Issuu

Geological Society of Australia

ABSTRACTS Number

SGGMP - Dunedin 2007

Specialist Group in Geochemistry, Mineralogy and Petrology Second bi-annual conference Dunedin, New Zealand 14th _ j9th October, 2007 Geological Society of Australia

Geological Society of Australia

ABSTRACTS Number

SGGMP - Dunedin 2007

Editor: Ian T Graham

Specialist Group in Geochemistry, Mineralogy and Petrology Second bi-annual conference Dunedin, Zealand 14th _ 19th October, 2007

Produced By: Ian Graham School of Biological, Earth and Environmental Sciences University of New South Wales Sydney, NSW 2052, Australia

October 2007

ISSN 0729 01IX

Purchase requests should be addressed to: Business Manager Geological Society of Australia Suite 706, 301 George Street Sydney, NSW 2000 © Geological Society of Australia Specialist Group in Geochemistry, Mineralogy and Petrology (SGGMP)

TABLE OF CONTENTS Page Simon D. Beams Jingera Rock Mesozoic alkaline complex: quartz and nepheline bearing syenite on the south-eastern Australian continental margin.

Stuart Mills, Bill Birch and Ian Grey A new uranyl phosphate mineral from the Lake Boga Granite, Victoria, Australia.

David M. Colchester, David E. Hibbs, Volker H. Hoppe, Peter Leverett 10 Adam R. McKinnon, James L. Sharpe, Peter Turner and Peter A. Williams Some things blue from The Great Australia Mine, Cloncurry District, Queensland. Alan F. Cooper and Lorraine A. Paterson Carbonatites from a lamprophyre dyke swarm. South Westland, New Zealand.

Nathan Daczko, Stephanie Carroll, Luke Milan and Geoffrey Clarke Age constraint and new field relationships of high-pressure mafic granulites of Fiordland, New Zealand.

Matthew C. De Paoli, G.L. Clarke and K.A. Klepeis The metamorphic evolution of the root to a magmatic arc: coronas and shear zones in the Western Fiordland Orthogneiss, Doubtful Sound, New Zealand.

John Foden, M. Hand, C. Lawley and M. Burdett Granite formation in extensional accretionary orogens: the generation of maximum I-S type granite geochemical distinction.

Ian T. Graham, Lee Spencer, Gregory Yaxley and Larry Barron The use of zircon in diamond exploration - a preliminary case study from the Cempaka deposit, SE Kalimantan, Indonesia.

Elena Hancock, Ernest Nickel, Michael Verrall and David Vaughan Gold-variscite mineralisation at Woodlands, Western Australia.

Heather Handley, Simon Turner, Ian Smith and Robert Stewart Timescales of magma genesis and differentiation at Lopevi Volcano, Vanuatu SW Pacific.

Ryan Ickert and Ian Williams 49 Correlated, in-situ analysis of U/Pb, and sHf in zircon from Siluro-Devonian granite in the eastem Lachlan Orogen: constraints on juvenile additions to the continental crust. Dushan Jugum, Richard Norris and Michael J. Palin 54 Permian to Jurassic Gondwanan accretion in New Zealand: constraints from basalt geochemistry and detrital zircon geochronology.

Luke A. Milan, Nathan R. Daczko, Ian Tumbull, Andrew Allibone and Geoffrey Clarke. A U-Pb and hafnium in-situ zircon investigation of orthogneiss and paragneiss units of Western Fiordland, New Zealand.

Nick Mortimer, Mike Palin and Rick Herzer 62 Miocene-Quaternary evolution of SW Pacific arcs and backarc basins: new ages and a choice of tectonic models. Janet R. Muhling, Birger Rasmussen and Ian R. Fletcher 65 Low-temperature metamorphic monazite: occurrence, chemistry and geochronology. Ian Nicholls, Adam Moir and Zarah Heyworth 71 Young trachyandesite lava domes of the Gisbome area, central Victoria, Australia: products of shallow or deeper crustal contamination of tholeiitic basaltic magmas? Tracy Rushmer, A. Getsinger and M.D. Jackson Origin of compositional variations in TTG-like magmas: an integrated experimental and numerical study.

James M. Scott, J.M. Palin and A.F. Cooper Dating of fabric development by U-Pb isotopes: case studies from Fiordland, New Zealand

F. Lin Sutherland, G. GiuUani, A.E. Fallick and G.B. Webb Oxygen isotopes in gem corundums, eastern Australia: further clues to their lithologic sources.

Andy Tulloch, Jahandar Ramezani, David Kimbrough, Kevin Faure and Andrew Allibone U-Pb geochronology of mid-Paleozoic granitic magmatism in New Zealand.

Simon Turner, Monica Handler and Ilya Bindeman New constraints from Tonga-Kermadec on the origin of O-Hf-Os isotopic signatures in oceanic arc lavas.

Jingera Rock Mesozoic alkaline complex: quartz and nepheline bearing syenite on the south-eastern Australian continental margin.

Simon D. Beams Terra Search Pty Ltd, PO BOX 981, Castletown, Hyde Park, Qld, Australia, 4812.

Jingera Rock is a 300 m high rock face, forming a prominent landmark in the coastal ranges, 15 km west of Eden, NSW, on the continental margin of south eastern-Australia (Fig. 1). The Mesozoic alkaline nature of syenites and associated rocks within the Jingera complex was not recognized until the mid 1970's. Up until that time, the rock itself and the range to the south simply appeared on published geological maps as Devonian granite.

Rock units of this complex intrude slates and sandstones of probable Ordovician age, together with granites of the Early Devonian Bega Batholith. Pyroxene homfels assemblages developed in the country rocks indicate P-T conditions of intrusion for the syenite complex of less than 1 kb and 700-735°C. Sheet silicates have been broken down by contact metamorphism: muscovite in peraluminous granite converted to andalusite-bearing and biotite in metaluminous granite converted to hypersthene-bearing aggregates. Anhydrous clinopyroxene assemblages also occur in the contact aureole. The interpreted high level of intrusion is supported by field evidence such as miarolitic cavities, porphyryritic and quench textures and intrusive breccia.

SGGMP - Dunedin 2007

MIcrosyenite Intrusive Breccia Biotite Nepheline Monzonite Nepheline Syenite Fayalite +/Quartz Syenite

JINGERA ROCK

Simplified Geology of the Jingera Rock Syenite Complex South East NSW Figure 1

Dr. 1. McDougall (written comm., 1978) reports a mean age from K-Ar determinations on biotite and hornblende of 168.2 ± 2.5 Ma.

This Jurassic age corresponds to a time of

continental break-up.

Although the Jingera Syenite Complex is one of several examples of Mesozoic alkaline magmatism which occur along the continental margin of eastem Australia, the following features make it unique: •

the large body of nepheline syenite. the intimate association of silica saturated and undersaturated rocks together with both peraluminous and peralkaline types. the presence of a fragmental monzonitic unit of probable intrusive breccia origin which contains an array of clasts and megacrysts that have the potential to allow the petrogenetic and magmatic history of the Jingera Complex to be unraveled.

SGGMP - Dunedin 2007

Syenite is the predominant iithology within the complex: both quartz- and feldspathoidbearing varieties are present. Fayalite-hedenbergite syenite is generally quartz-bearing and forms a narrow (5 to 50 m wide) outer rim. It also occurs as a series of discontinuous screens within the nepheline-bearing syenite (Fig. 1).

The bulk of the complex consists of nepheline syenite which is a coarse-grained rock dominated by microperthitic feldspar, nepheline and prominent black amphibole which may appear as bladed crystals up to 100 mm long. Large grains of nepheline and sodalite are interstitial. Hedenbergite-aegerine augite is generally present along with the following accessory minerals: titanomagnetite, apatite, allanite, titanite, and zircon, and rarer Zr minerals such as baddeleyite and eudialyte. Nepheline syenite intrudes quartz syenite.

Biotite nepheline monzonite is another coarse-grained silica undersaturated unit with large flakes of biotite and complexly zoned, well-formed crystals of plagioclase having prominent oscillatory zoning and surrounded by a rim of alkali feldspar. Nepheline, sodalite and alkali feldspar are interstitial.

Matrix supported intrusive breccia form small bodies that cut the fayalite ± quartz syenite but are intruded by nepheline syenite. The intrusive matrix of a just silica under-saturated microsyenite consists of alkali feldspar, hedenbergitic pyroxene, minor amphibole and biotite. Abundant sub-angular to sub-rounded clasts (5-100 mm) are of various lithologies including coarse and fine fayalite-hedenbergite syenite, quartz trachyte and country rock homfels. Other inclusions include single megacrysts and polycrystalline aggregates. Mafic aggregates consist of titanaugite, kaersutite, titanomagnetite, and minor ilmenite. Plagioclase is abundant as complexly zoned single grains which nominally impart a monzonitic composition to the breccia. Anorthoclase also occurs as large crystals. Some rare inclusions are altered olivine, surrounded by prominent reaction haloes of amphibole and biotite. Other breccias have quartz-bearing syenite and sodalite phonolite intrusive matrix.

Most of the inclusion-bearing intrusive matrix breccias plot close to the boundary of silica saturation in the system Si02 - NaAlSi04 - KAlSi04 (Fig. 2). The clast assemblage entrained

SGGMP - Dunedin 2007

in the microsyenite matrix contains the various rock types that the intrusion has encountered during its probable rapid and explosive ascent, from where it was generated in the lower crust to its current position in the upper crust. The larger clasts are mainly rock types present within, or marginal to the complex. However, the megacrysts and crystal aggregates contain minerals out of equilibrium with the pressure-temperature environment of the upper crust: e.g. reaction rims are prominent, magnetite, hercynitic spinel and pyroxene inclusions are aluminous. Overall the inclusion assemblage of aluminous clinopyroxene + sodic plagioclase + Mg-Al spinel + titanomagnetite =t olivine + minor hypersthene + minor sanidine (anorthoclase) has a gabbroic composition. The interpretation is that the inclusions are derived from a transitional basalt composition at about 10 kb.

Both silica oversaturated and undersaturated dyke rocks occur throughout the complex and also cut surrounding country rocks.

The most evolved dykes are peralkaline phonolite,

sodalite phonolite and quartz trachyte. Fluorite-rich trachyte and syenite are the latest dykes in the Jingera Rock area cutting all other rock types.

A broad spectrum of compositions from silica undersaturated to silica oversaturated and from peralkaline to metaluminous rocks is present in the Jingera Syenite Complex. In spite of this range in composition, all the rocks of the complex have a distinctive chemical character which is unique in southeast Australia.

These distinctive features include high Na, total

alkalis, La, Ce, Nd and Ga, most have high Zr and high to extremely high Nb.

The fact that the Jingera rocks form a distinctive geochemical suite throughout the whole range of compositions indicates a direct genetic relationship between saturated and undersaturated syenite. However, there is a the high degree of scatter on chemical variation diagrams and a concomitant absence of differentiation trends particularly within the syenites, of ratios normally regarded as indices of feldspar fractionation (e.g. K/Rb, Rb/Sr, Ba/Rb, Ba/K). In addition, there is an absence of mafic rocks (less than 54% Si02), indicating that basalt fractionation, at least at low pressures, is unlikely. The different rock types present cannot be related by crystal fractionation of a single parent. The composition of inclusions contained within both quartz and nepheline normative microsyenite preclude the derivation of the complex from typical mantle material.

SGGMP - Dunedin 2007

Available data indicates that the source of the syenites had transitional basalt chemistry. Partial melting of such basalt at around 10 kb and 1100°C can produce a series of melts with the observed range of silica saturation, especially if small amounts of F and CI are also available. There is good petrographic evidence that F is high in some rocks of the complex (e.g. fluorite is relatively abundant in certain saturated syenites containing anhydrous mafic assemblages). Cl-rich minerals (e.g. sodalite) occur in the nepheline syenites and are abundant in the phonolite dykes. High F and CI are probably responsible for the high abundances of large highly charged cations (e.g. Zr, Nb, Ga, Y, REE) as well as certain transition elements (e.g. Zn) which are more readily retained in syenite magma.

In summary, it is concluded that at least in the case of the Jingera Syenite Complex, the intimate association of oversaturated and undersaturated compositions is the result of slight differences in the composition of the transitional basalt source rock and small differences in the conditions of partial melting at about 10 kb (i.e. close to the crust-mantle boundary). There has been no fractionation across the thermal barrier. Low pressure fractionation is away from the feldspar join in the system Si02 -NaAlSi04 - KAlSi04 and leads to more extreme compositions: one oversaturated and the other undersaturated (Fig. 2). Those rocks plotting close to the oversaturated minimum contain no inclusions and their chemical compositions (e.g. very low Ba, and Sr, and high Rb and Ga) can be explained by feldspar fractionation. Likewise, phonoHte dyke rocks that correspond in composition to the minimum in the undersaturated part of the system Si02- NaAlSi04 - KAlSi04, are also devoid of inclusions and display chemical and mineralogical patterns indicative of feldspar fractionation and peralkaline compositions.

SGGMP - Dunedin 2007

NaAlSi308

KAlSi04

NaAlSi04

NaAISi308

KAiSiSOS

Figure 2. Rocks from the Jingera Rock Syenite Complex plotted in terms of the System Si02 -NaAlSi04 - KAlSi04. ("Petrogenys Residual System"). :A: all samples; B: syenites and monzonites only on expanded part of diagram across albite - orthoclase join; C: Inclusionbearing intrusive matrix breccia samples only; Filled symbols: Quartz-bearing fayalite syenite and quartz trachyte; Unfilled symbols: feldspathoidal syenite and phonolite; Filled triangle: inclusion-bearing intrusive matrix breccia.

SGGMP - Dunedin 2007

A new uranyl phosphate mineral from the Lake Boga Granite, Victoria, Australia

Stuart Mills^ Bill Birch^ and Ian Grey^

^ Department of Earth and Ocean Sciences, University of British Columbia, Vancouver, Canada ^ Museum Victoria, Melbourne, Victoria, Australia ^ CSIRO Minerals, Clayton, Victoria, Australia

A new uranyl phosphate mineral, ideally CaNaFe^^2H(U02)2(P04)4(0H)2(H20)8, occurs in the Late Devonian Lake Boga Granite, northern Victoria, Australia. It is one of five secondary uranium species found in the weathering zone of the granite, whose only exposure is a large operating quarry 10 km SSW of Lake Boga township. Torbemite, saleeite, metanatroautunite ('sodium autunite'), ulrichite (new species in 1988) and the new mineral all occur as millimetric-scale crystals in miarolitic cavities and on joint planes encrusted with iron oxy-hydroxides. The new mineral occurs as bright lemon-yellow transparent prismatic crystals up to 0.4 mm across, on a matrix of microcline, albite, smoky quartz and muscovite. Other secondary minerals associated with it are torbemite and the sodium analogue of meurigite. The crystals have a vitreous lustre and a pale yellow streak. Mohs hardness is about 3. A crystal- structure study revealed pairs of edge-shared UO7 pentagonal bipyramids that are inter-linked via comer- and edge-sharing with PO4 tetrahedra, to form chains parallel to the c-axis. These (U02)2(P04)4 chains are cross-linked via comer-sharing between the PO4 tetrahedra and Fe04(0H)2 octahedra. The Na^ and Ca^^ cations, and 4 water molecules occupy eight-sided channels along [010]. The remaining water molecules occupy large tensided channels directed along [001]. This stmcture is unique amongst those of U^^ minerals and synthetic uranyl compounds (Mills et al in press

SGGMP - Dunedin 2007

8 The Lake Boga Granite contains up to 65 ppm uranium, with most contained within primary accessory uraninite and monazite (up to 8 wt% U) (Mills et al in press^). Late-stage hydrothermal fluorapatite crystals also contain uranium, as thin U-enriched zones parallel to crystal growth pattems, and as sporadic uraninite inclusions. Weathering of uraninite and fluorapatite within the upper few metres of the granite resulted in the groundwater becoming locally enriched in U, P and Ca. The presence of patches of magmatic chalcopyrite in the same zone had the effects of lowering pH and providing Cu and Fe for groundwater (Acero et al, 2007). The new mineral and the other, more common uranyl phosphates in the Lake Boga Granite have crystallised from these groundwaters. Because U is rapidly fixed in uranyl phosphates (Murakami et al, 1991 \ Jerden and Sinha, 2003), it is likely that crystallisation has taken place only short distances from the sites of oxidation. Attempts at U-Th dating of the uranyl phosphates from the Lake Boga quarry have yielded apparent ages from 115 to >550 ka (Maas et al, 2006), although these need to be assessed against variable crystal growth mechanisms and possible loss/gain of U during repeated groundwater residence. The pristine appearance of the uranyl phosphate crystals does not suggest these factors have been major influences on the apparent ages. It is likely therefore that the uranyl phosphates occurring in the Lake Boga Granite have acted to sequester uranium released during weathering of the granite over time scales of several hundred thousand years. This aspect of the uranium geochemistry and geochronology of the Lake Boga Granite is under further study.

REFERENCES Acero, P., Cama, J. And Ayora, C., 2007. Kinetics of chalcopyrite dissolution at pH3. European Joumal of Mineralogy 19: 173-182. Jerden, J.L. and Sinha, A.K., 2003. Phosphate based immobilization of uranium in an oxidising bedrock aquifer. Applied Geochemistry 18: 823-843. Maas, R., Mills, S.J., Birch, W.D. and Hellstrom, J., 2006. Secondary U-phosphates — possible links to Quatemary pluvial periods in SE Australia. Geochimica et Cosmochimica Acta 70: 379. Mills, S.J., Birch, W.D., Maas, R., Phillips, D. and Plimer, LR. (in press^). Lake Boga Granite, northwestem Victoria: mineralogy, geochemistry and geochronology. Australian Joumal of Earth Sciences. SGGMP - Dunedin 2007

Mills, S.J., Birch, W.D., Kolitsch, U., Mumme, W.G. and Grey, I.E. (in press^). , CaNaFe^^H(U02)2(P04)4(0H)2(H20)8, a new uranyl phosphate with a unique crystal structure from Victoria, Australia. American Mineralogist. Murakami, T., Ohnuki, T., Isobe, H. and Sato, T., 1997. Mobility of uranium during weathering. American Mineralogist, 82: 888-899.

SGGMP - Dunedin 2007

Some things blue from The Great Australia Mine, Cloncurry District, Queensland

David M. Colchester\ David E. Hibbs^ Volker H. Hoppe^ Peter Leverett\ Adam R. McKinnon\ James L. Sharped Peter Tumer^ Peter A. Williams\

^ School of Natural Sciences, University of Western Sydney ^ School of Pharmacy, University of Sydney ^ School of Chemistry, University of Sydney ^ 46 Dalmor Ave. Ormond, Victoria

An ongoing study of oxide zone minerals, especially blue copper-bearing ones from the Great Australia Mine 2 km south of Cloncurry, Queensland, has resulted in the identification of a new mineral (cloncurryite) and the crystal chemical characterization of another (connellitebuttgenbachite). Other blue minerals occurring in the oxidized zone are cometite, gerhardite and libethenite.

The Great Australia Mine (20^^ 42' S, 140^ 30' E) was discovered in 1867 and has been mined intermittently since. It lies in the Mount Isa - Cloncurry Mineral Field situated in the Mount Isa inlier consisting of north-northwesterly trending Early to Middle Proterozoic rocks. These rocks consist of sequences of felsic volcanics and shales intruded by a number of mafic and granitic bodies of varying ages. To the west and south the inlier is overlain by younger Proterozoic and Palaeozoic rocks (Blake et al, 1986). The primary sulfide mineralogy consists of cobalt-bearing pyrite and chalcopyrite in a quartz - albite - actinolite - dolomite calcite gangue (Cannel and Davidson, 1998). The deposit is fauh controlled and related to the Williams-Naraku granite suite (Wybom, 1998).

SGGMP - Dunedin 2007

11 Cloncurryite The new mineral Cloncurryite, IMA 2005-060 (Colchester et. al, 2007) occurs as small, slender acicular crystals with a sky-blue colour. They occur in erinaceous clusters in silicified goethite-hematite gossan in a restricted zone near the junction of the Main and B Tangye loads on bench 6 of the open-cut and about 20 m from the surface. Conveniently, some of this material was dumped on the surface making it readily available for fossicking. Associated copper minerals include malachite, pseudomalachite, cuprite and native copper. Since the powder XRD pattern did not match any spectra in the ICDD data file, microprobe analyses were made and single crystal data collected. As a result of this experimental work the following data were collected.

The microprobe analysis gave a composition: CuO 10.29, VO2 8.32, AI2O3 23.63, Fe203 0.32, P2O5 32.54, F 4.34 with H2O (by difference) 22.4 giving an empirical formula based on 2 PO4 groups pfu and (F + OH) equal to 2 pfu of: [CU0.56(VO)0.44]ll00(Al2.02Fe0.02)l2.04(PO4)2Fi.00OHi.00*4.92H2O and a simplified formula of: CU0.56(VO)0.44Al2(PO4)2(F,OH).5H2O

The presence of the vanadyl ion (VO)^"" was confirmed during the single crystal X-ray structure analysis. The sheet-like structure of cloncurryite is produced by Al, phosphate, fluoride and hydroxide ions and coordinated water molecules is comparable to that in nevadaite. Both aluminium ions are octahedrally coordinated; one is bonded to two trans water molecules, two cis phosphate oxygen atoms and two disordered cis (F, OH) ions and the other to four phosphate oxygen atoms and two trans disordered (F, OH) ions. A104(F,0H)2 octahedra are linked in chains parallel to the b axis by (F, OH) bridges and phosphate oxygen atoms, and the latter also serve to link the chains into sheets parallel to the ah plane. The sheets are further linked in the c direction via coordination of phosphate oxygen atoms to Cu^^ and VO^^. Cu (1) is coordinated in a square planar fashion to two phosphate oxygen atoms (cw-O (3) and O (7)) and two water molecules (cw-0W3 and 0W4) in folly occupied sites. A water molecule completes a Jahn-Teller distorted square pyramidal coordination sphere. The site occupancy factor (sof) for Cu (1) is 0.53. Cu (2), which is

SGGMP - Dunedin 2007

12 present only in very small amounts, is similarly coordinated to the same atoms, but in a related face-sharing position down the a axis and assumes a similar square pyramidal geometry. VO^"^ is coordinated to the same four oxygen atoms as Cu (1) and site occupancy factors of the V atom and the vanadyl O atom are equal at 0.44. The Cu-0W5 vector is opposite in direction to the V=0 vector. It should be noted that only one of the Cu (1), Cu (2) and V sites is occupied at any one time.

Nevadaite and Cloncurryite Our initial submission to The IMA CNMMN for new mineral status was rejected on the grounds of its similarity with the mineral nevadaite which has the formula: (Cu'^n,Al, V'^)6[Al8(P04)8F8](0H)2(H20)22 Both minerals have similar optical properties, and unit cell dimensions (Cooper et. al, 2004) but nevadaite is orthorhombic (cf monoclinic for cloncurryite), the vanadium is present as V

(cf V^'" as VO "^in cloncurryite) and the copper is present as Cu^^.(cf Cu"^ in cloncurryite).

However, the CNMMN wanted to see if a crystal structure determination of cloncurryite based on an orthorhombic lattice could be resolved.

Connellite Blue acicular crystals of nitrate-rich connellite also occur in the oxidized zone. The crystal chemistry of this sequence was characterized by Hibbs et al (2003). The single crystal structure of a specimen of connellite from the Great Australia mine was determined using synchrotron

radiation.

The

derived

formula

for

the

crystal

concerned

Cu36Cl7.82(SO4)0.52(NO3)0.48(OH)62.66*5.12H2O. Given the nearly equal amounts of sulfate and nitrate in the lattice, it is likely that both connellite and buttgenbachite occur in the deposit, which is noted, among other things, for its secondary copper nitrate mineralization. Both gerhardite, Cu2N03(0H)3, and likasite, Cu3N03(0H)5*2H20, are knovm from the deposit. The nitrate of the supergene sequence has been shown to be biogenic in origin (Melchiorre et a l , 2006).

SGGMP - Dunedin 2007

13 REFERENCES Blake, D.H., Stewart, A.S., Bain, J.H.C., Page, R.W., Wybom, L.A. and Etheridge, M.A., 1986. Project 2B.02 Metallogeny and crustal evolution in the Mt. Isa Province. Yearbook of the Bureau of Mineral Resources, Geology and Geophysics. Australian Government Publishing Service. Cannell, J. and Davidson, G.J., 1998. A carbonate-dominated copper cobalt breccia-vein system at the Great Australia deposit. Mount Isa eastern succession. Economic Geology 93: 1406-1421. Colchester, D.M., Leverett, P., McKinnon, A.R., Sharpe, J.L., Williams, P.A., Hibbs, D.E., Turner, P. and Hoppe, V.H., 2007. Cloncurryite, Cu0.56(VO)0.44Al2(PO4)2(F,OH)*5H2O a new mineral from the Great Australia mine, Cloncurry, Queensland, Australia, and its relationship to nevadaite. Australian Journal of Mineralogy 13: 5-18. Cooper, M.A., Hawthorne, F.C. Roberts, A.C. Foord, E.E., Erd, F.C., Evans, H.T.Jr., and Jensen, F.C., 2004. Nevadaite, (Cu^^D, Al, V^^)6[Al8(P04)8F8](0H)2(H20)22, a new phosphate mineral species from Gold Quarry mine Carlin, Eureka County, Nevada: description and crystal structure. Canadian Mineralogist 42: 741-752. Hibbs, D.E., Leverett, P., WiUiams, P.A., 2003. Connellite-buttgenbachite from the Great Australia mine, Cloncurry: a structural formula. Australian Journal of Mineralogy 9: 39-42. Melchiorre, E.B., Williams, P.A., Rose, T.P., and Talyn, B.C., 2006. Biogenic nitrogen from termite mounds and the origin of gerhardite at the Great Australia mine, Cloncurry, Queensland, Australia. Canadian Mineralogist 44: 1447-1455. Wybom, L., 1998. Younger ca 1500 Ma granites of the Williams-Naraku Batholiths, Cloncurry district, eastern Mount Isa Inlier: geochemistry, origin, metallogenic significance and exploration indicators. Australian Journal of Earth Sciences 45: 397411.

SGGMP - Dunedin 2007

Carbonatites from a lamprophyre dyke swarm, South Westland, New Zealand

Alan F. Cooper, Lorraine A. Paterson Geology Department, University of Otago, P.O. Box 56, Dunedin, New Zealand

In the Southern Alps of northwestern Otago and south Westland, the Haast Schist basement is intruded by a regional swarm of predominantly lamprophyre dykes that extends approximately 100 km from the vicinity of Wanaka to the headwaters of the Paringa River (Fig. 1). This igneous province, termed the Alpine Dyke Swarm by Cooper (1986), has a possible outlier a further 50 km to the south at Nevis Bluff (Hutton, 1940). Although the swarm has a NNE trend, individual dykes have preferred east-west orientations that were interpreted by Cooper et al (1987) as due to intrusion along Riedel shears associated with the transtensional propagation of the Alpine Fault plate boundary through the South Island, New Zealand.

U-Pb and Rb-Sr ages (Barreiro and Cooper, 1987; Cooper et al, 1987) and K-Ar studies on samples distant from the plate boundary and therefore less affected by recent uplift (Adams and Cooper, 1996), gave ages that are interpreted as representing intrusion in the late Oligocene-early Miocene (32-20 Ma). The Nevis Bluff limburgite has recently been dated by Ar-Ar methods at 20.7 ± 0.4 Ma (Hoemle et al, 2006).

Intrusion is most intense in the Haast-Burke River area of south Westland, where the swarm comprises feldspar-free ultramafic lamprophyres (aillikite and damtjemite) and plagioclasebearing (camptonite) lamprophyres, tinguaite (porphyritic phonolite), trachytes and carbonatites.

Lamprophyres are characteristically ocellar, with rounded, leucocratic ocelli filled with calcite, analcime, alkali feldspar and minor ferromagnesian phases. At high erosion levels exposed in the southern part of the swarm, the glassy groundmass of chilled margins and SGGMP - Dunedin 2007

15 crystalline ocellar infillings and segregation veinlets in lamprophyres have similar evolved phonolitic compositions (Cooper, 1979).

Carbonatite intrusions make up less than 1% of the total volume of magma injected into the schists of the Haast River area (Cooper, 1986), with the host metasomatically altered by fenitization to proximal aegirine-albite and distal riebeckite/Mg-arfvedsonite-bearing schists. Fenitizing fluids are generally rich in Na, Mn, Ba, Sr, Nb, and CO2, with CI, F and Li suggested by the presence of baotite and tainiolite.

Textures of the Haast carbonatites are commonly allotriomorphic granular with banding defined by variations in grain size or segregations of contrasting mineralogy, including carbonate, silicate and sulfide phases. Many carbonatites are inequigranular, with coarse rounded porphyroclasts enclosed in a fine-grained, mylonitised groundmass.

Carbonatites range from cream coloured varieties containing combinations of calcite, dolomite, norsethite (Cooper, 1971), strontianite, daqingshanite-(Ce) (Cooper, 1986), BaREE-'Na carbonate, carbocemaite, barytocalcite/paralstonite, an unnamed Ca-Sr phase, and at least three as yet unidentified Ca-Ba-Sr-iJ^^ minerals. Brown weathering varieties are dominated by ankerite and/or siderite. Mineral assemblages commonly contain fine-grained aggregates of multiple carbonate combinations. Silicate phases include albite and rarely Kfeldspar, aegirine, Na-amphibole, and rare baotite (Cooper, 1996) and stilpnomelane (Paterson, 1993). Other accessory minerals include apatite, monazite, sporadic barite, hematite, Nb-rutile, thorite, rare fergusonite, and ubiquitous pyrite, pyrrhotite, chalcopyrite, sphalerite, and galena. Modal analyses calculated from whole rock (XRF) compositions and microprobe compositions of individual minerals yield four broad and overlapping groups that comprise: calcite-, dolomite-ankerite-, siderite-, and norsethite-rich varieties (Fig. 1).

SGOIP - D u n e d i n

2007

Mineral Compositions FeCO Carbonat^te Type • calclte O dolomite - ankarite 4? siderite *if norsethite

Rock Compositions FeCO ^MnCO. Carbonatitfi Type o c^cite # dolomite - ankerite siderlte T norsethite

CaCO CaCO

MgCO

Fig 1. a. Ca-Mg-Fe carbonates from calcite, dolomite-ankerite, siderite, and norsethite carbonatite types in the Alpine Dyke Swarm, Haast River area. Phase fields are from Anovitz and Essene (1987) and Rosenberg (1967) (shaded), b. Carbonatite whole rock compositions expressed in terms of CaCOs-FeCOs + MnCOs-MgCOs. CC-calcio-carbonatite, FCCferruginous calcio-carbonatite, FC-ferro-carbonatite and MC-magnesio-carbonatite represent modified classification fields from Gittins and Harmer (1997), modified after Woolley and Kempe (1989).

A variety of geothermometric estimates (often unreliable in carbonatite systems) and constraints from mineralogical assemblages suggest temperatures of 600''C to 250''C spanning the spectrum from magmatic to carbothermal processes. Mineralogical textures indicate, in places, progressive interaction of primary liquidus? minerals with a Ba-Sr-7?^^rich medium. REE concentrations are largely controlled by the presence of monazite that shows a range in compositions from the straight-line, log-linear LREE pattern seen in most classical carbonatites to lifE'^-depleted, MiJE^-enriched, convex upward pattems that may represent a primary feature of late-stage carbonatite varieties, or could originate by lowtemperature interaction with carbothermal fluids.

SGGMP - Dunedin 2007

17 Dykes from the Haast River segment of the swarm, ranging in composition from ultramafic lamprophyre to norsethite-rich carbonatite show very primitive ^'Sr/^^Sr ratios, ranging from 0.70270 to 0.70348. '"'^Nd/'^'^Nd ratios show a similar restricted range from 0.51282 to 0.51291. Pb isotope data indicate radiogenic lead, with 19.18 - 20.60, and ^^^b/^^b 39.34 - 40.22. In a s r - S

15.64 -15.71, diagram, samples plot to the left

of the mantle array defined by MORE and OIB in the depleted quadrant, indicating derivation from a mantle source that had experienced long-term depletion in the LREE and in Rb/Sr. Isotope ratios from the Dunedin Volcanic Group (Coombs et al, 1986; Price et al, 2003) are indistinguishable from the Alpine Dyke Swarm analyses, as are analyses of the limburgite from Nevis Bluff and a lamprophyre from the southern part of the swarm (Hoemle etal.,2006).

Bell and Tilton (2002) showed that so-called Reference Carbonatites (Tilton et al., 1998) of which the Alpine Dyke Swarm suite forms a part, plotted on ^^Sr/^Sr and '''^Nd/'^'^Nd versus 206p|^/204p^ diagrams as a mixture of HIMU and FOZO. Using redefined compositions for FOZO (Stracke et al, 2005), magmas of the Alpine Dyke Swarm can be explained by melting of the single mantle component, FOZO.

Using the criteria of Eggler (1989), the Haast carbonatites are too fractionated to represent direct primitive melts from this mantie source. Instead, several evolutionary paths are possible: (a) separation by fractional crystallisation from a carbonated silicate rock parent (experimental process described by Lee and Wyllie 1998, mechanism advocated to describe natural occurrences by Bell and Rukhlov 2004; Downes et al, 2005), or (b) separation by liquid immiscibility from a carbonated silicate parent.

While the generation of tinguaites by fractional crystallisation of lamprophyres is well established (Cooper 1979, 1986), the origin of the carbonatites is uncertain. Volumetrically, calcite does not increase progressively from lamprophyre to carbonatite, and an origin of carbonatite by fractional crystallisation is not realistic. An origin of carbonatite by liquid immiscibility has been suggested, with the parent carbonated silicate magma being either ultramafic lamprophyre (Rock, 1986), feldspathic lamprophyre (Ferguson and Currie, 1971)

SGGMP - Dunedin 2007

18 or undersaturated nepheline syenite/phonolite (Andersen, 1988).

On the basis of the partition coefficient data, mineralogical evidence (particularly the aegirine-rich pyroxene in late-stage phonolitic liquids in lamprophyre groundmasses, in tinguaites and in proximal fenites), and the close field association of carbonatite and tinguaite, it is suggested that the carbonatites and a (potential fenitising) fluid phase have separated immiscibly from a tinguaite parental magma.

Subsequent magmatic differentiation of the carbonatite and interaction with carbothermal fluids to low^ temperatures is capable of generating the spectrum of carbonatite compositions observed.

REFERENCES Adams, C.J., and Cooper, A.F., 1996. K-Ar age of a lamprophyre dike sv^arm near Lake Wanaka, west Otago, South Island, New Zealand. New Zealand Journal of Geology and Geophysics 39: 17-23. Andersen, T., 1988. Evolution of peralkaline calcite carbonatite magma in the Fen complex, southeast Norway. Lithos 2: 9-112. Anovitz, L.M., Essene, E.J., 1987. Phase equilibria in the system CaCOs-MgCOs-FeCOs. Journal of Petrology 28: 389-414. Barreiro, B.A., and Cooper, A.F., 1987. A Sr, Nd, Pb isotope study of alkaline lamprophyres and related rocks from Westland and Otago, South Island, New Zealand. Geological Society of America Special Paper 215: 115-125. Bell, K., and Rukhlov, A.S., 2004. Carbonatites from the Kola Alkaline Province: origin, evolution and source characteristics. In: Phoscorites and carbonatites from Mantle to Mine (Wall, F., and Zaitsev, A.N., eds.). Mineralogical Society Serial 10: 433-468. Bell, K.D., and Tihon, G.R., 2002. Probing the mantle: the story from carbonatites. EOS Transactions 83: 273-277. Coombs, D.S., Cas, R., Kawachi, Y., Landis, C.A., McDonough, W.F. and Reay, A., 1986. Cenozoic volcanism in north, east and central Otago. In: Late Cenozoic Volcanism in

SGGMP - Dunedin 2007

19 New Zealand (Smith, I.E.M., ed.). Royal Society of New Zealand Bulletin 23: 278312. Cooper, A.F., 1971. Carbonatites and fenitization associated with a lamprophyre dike-swarm intrusive into schists of the New Zealand geosyncline. Geological Society of America Bulletin 82: 1327-1340. Cooper, A.F., 1979. Petrology of ocellar lamprophyres from western Otago, New Zealand. Journal of Petrology 20: 139-163. Cooper, A.F., 1986. A carbonatitic lamprophyre dike swarm from the Southern Alps, Otago and Westland. In: Late Cenozoic volcanism in New Zealand (Smith, I.E.M., ed.). Royal Society of New Zealand Bulletin 23: 313-336. Cooper, A.F., 1996. Nb-rich baotite in carbonatites and fenites at Haast River, New Zealand. Mineralogical Magazine 60: 473-482. Cooper, A.F., Barreiro, B.A., Kimbrough, D.L., and Mattinson, J.M., 1987. Lamprophyre dike intrusion and the age of the Alpine Fault, New Zealand. Geology 15: 941-944. Downes, H., Balaganskaya, E., Beard, A., Liferovich, R., and Demaiffe, D., 2005. Petrogenetic processes in the ultramafic, alkaline and carbonatitic magmatism in the Kola Alkaline Province: a review. Lithos 85: 48-75. Ferguson, J., and Currie, K.L., 1971. Evidence of liquid immiscibility in alkaline ultrabasic dikes at Callander Bay, Ontario. Journal of Petrology 12: 561-585. Gittins, J., and Harmer, R.E., 1997. What is ferrocarbonatite? A revised classification. Journal of African Earth Sciences 25: 159-168. Hutton, C.O., 1940. Limburgite from Nevis Bluff, Kawarau Gorge, Central Otago. Transactions of the Royal Society of New Zealand 73: 58-67. Lee, W-J., and Wyllie, P.J., 1998. Processes of crustal carbonatite formation by liquid immiscibility and differentiation, elucidated by model systems. Journal of Petrology 39: 2005-2014. Paterson, L.A., 1993. A study of carbonatites and associated fenitization at Haast River, south Westland, New Zealand. Unpublished PhD thesis. University of Otago, Dunedin, New Zealand. Price, R.C., Cooper, A.F., Woodhead, J.D., and Cartwright, 1., 2003. Phonolitic diatremes

SGGMP - Dunedin 2007

20 within the Dunedin Volcano, South Island, New Zealand. Journal of Petrology 44: 2053-2080. Rock, N.M.S., 1986. The nature and origin of ultramafic lamprophyres: alnoites and allied rocks. Journal of Petrology 27: 155-196. Rosenberg, P.E., 1967. SubsoUdus relations in the system CaCOs-MgCOs-FeCOa between 350° and 550°C. American Mineralogist 52: 787-796. Stracke, A., Hofmann, A.W., and Hart, S.R., 2005. FOZO, HIMU, and the rest of the mantle zoo. Geochemistry, Geophysics, and Geosystems 6: Q05007, doi: 10.1029/2004GC000824. Tilton, G.R., Bryce, J.G., and Mateen, A., 1998. Pb-Sr-Nd isotope data from 30 and 300 Ma collision zone carbonatites in northwest Pakistan. Journal of Petrology 39: 1865-1874. Woolley, A.R., and Kempe, D.R.C., 1989. Carbonatites: nomenclature, average composition and element distribution. In: Carbonatites: Genesis and Evolution (Bell, K. ed.). UnwinHyman, London: 1-14.

SGGMP - Dunedin 2007

Age constraint and new field relationships of high-pressure mafic granulites of Fiordland, New Zealand

Nathan Daczko', Stephanie Carroll', Luke Milan' and Geoffrey Clarke^

^ GEMOC ARC National Key Centre, Department of Earth and Planetary Sciences, Macquarie University, NSW 2109, Australia (ndaczko@.els.ma.edu.au. scarrol]@.els.mq.edu.au. lmi]an@,els.mq.edu.au) ^ School of Geosciences, University of Sydney, NSW 2006, Australia (geoffc@,mail.usvd.edu.au):

Introduction Recent controversy surrounding interpretations of high-P granulites of Fiordland, New Zealand centre upon the driving mechanism for the transition from two-pyroxene hornblende granulite to garnet granulite. Blattner (2005) explains the development of the garnet granulite textures as involving metasomatic dehydration in response to (i) changes in P, T, chemical activity or stress, and (ii) physio-chemical field gradients driven by low-aH20 anatectic melts, at an unknown distance (possibly fifty to hundreds of metres) from the incipient garnet granulite textures. Clarke et al. (2005) reaffirms a model presented by Daczko et al. (2001) that interprets dehydration and transition to garnet granulite driven by volatile scavenging next to a migrating trondhjemitic liquid, sourced from the partial melting of nearby bodies of dioritic gneiss. This contribution presents new petrographic and geochronological data and field observations from a previously unrecognised style of garnet granulite texture at Lake Grave, near Sutherland Sound.

Regional geology The geology of the south island of New Zealand is divided into three domains. A belt of rocks referred to as the Median Tectonic Zone or Median Batholith separates Eastern and Western

Provinces.

The

Western

Province

contains

extensive

Lower

Palaeozoic

metasediments that are cut by Devonian and Carboniferous granitoids (Fig. 1). Rocks of the SGGMP - Dunedin 2007

22 Median Tectonic Zone and the Western Province were intruded by plutons of the 126-105 Ma Western Fiordland Orthogneiss / Separation Point Suite. The textures described in this study are in samples of the Western Fiordland Orthogneiss.

New Zealand Fiordland

Fiordland ^^^ m

J '•-•7 WFB x , r

m^ EFB

/'M/

//j

v, \\,

Alpine

iiftll f

* SWFB

Figure 1: Fiordland is located in the SW of the south island of New Zealand, east of the Alpine Fault. The Fiordland block comprises the (i) Western Fiordland Belt (WFB), including the Palaeozoic Tuhua Sequence and Cretaceous Western Fiordland Orthogneiss; (ii) Eastem Fiordland Belt (EFB), including the Median Tectonic Zone and Arthur River Complex; and (ii) Southwest Fiordland Block, including low-grade Palaeozoic units. Lake Grave is located -20 km south of Milford Sound.

New field relationships The Western Fiordland Orthogneiss at Lake Grave is uniform dioritic gneiss that displays granoblastic gneissose texture (SI) defined by irregular mafic aggregates of clinopyroxene, orthopyroxene and pargasite with or without Fe-Ti oxide set in a coarser-grained plagioclase matrix with minor potassium feldspar and quartz. SI is cut by moderately to steeply dipping, 5-20 mm thick trondhjemitic dykes that have unusual gamet granulite textures developed adjacent to them (Fig. 2). The gamet granulite texture is asymmetric with respect to the dyke and may show a semi-circular cross section, commonly only on one side of the dyke. We suggest that the gamet granulite textures at Lake Grave represent a rarely preserved earlier stage in the development of the common planar gamet reaction zones preserved across Fiordland. In addition, the identification of biotite-clinopyroxene symplectites at the gamet granulite front suggests that the chemical equations published for the transition from twopyroxene homblende granulite to gamet granulite (e.g. Daczko et al, 2001; Blattner, 2005) may not necessarily indicate the actual reaction, but rather the net change that took place

SGGMP - Dunedin 2007

23 during the transition.

Figure 2: Very narrow trondhjemitic dyke and associated garnet reaction zone (GRZ) or garnet granulite texture.

New age constraint Zircon is a common accessory mineral in the dykes. The zircon grains are commonly irregular and display grain shapes that mould adjacent grains (Fig. 3a). We interpret the grains to have grown in a crystal mush of cumulate phases late during the crystallisation of the dyke. U-Pb analysis of 27 separated zircon grains yields an age of 115 ± 0.74 Ma (MSWD = 0.83, probability = 0.71; Fig. 3b). U - P B Data - Concordia:

LG40S1'-V

134o o. .. . 1300 6

1260

v \

—

0 102P 0 0.02

0.06

0.10

0.14

0.1

Figure 3a (left): Irregular-shaped zircon grains (Zrc) with plagioclase feldspar (Flag), quartz (Qtz), apatite (Ap), scapolite (Sep), ilmenite (Ilm), pargasite (Prg), orthopyroxene (Opx) and clinopyroxene (Cpx). Figure 3b (right): U-Pb concordia diagram. The 27 analyses used in the age determination are shown as ftill lines. The 4 analyses excluded are shown as dashed lines.

SGGMP - Dunedin 2007

24 Discussion and conclusions The dykes, here described from Lake Grave, are different to the plagioclase feldspar-rich examples common across Fiordland. Their composition and textures are consistent with an igneous origin, despite the thin nature of the dykes (<10 to 20 mm across). We suggest that these dykes at Lake Grave represent a rarely preserved earlier stage in the development of the common "anorthositic" veins and dykes preserved across Fiordland.

The minor metasomatism associated with the garnet granulite texture involves, on average, loss of -0.4 wt% H2O, --1.3 wt% Na20 and possibly Si02 (Blattner, 2005). It remains unknown where this material goes to, however, material transport within the crust is usually accomplished either by magmatic or fluid flow. The question then arises if the amount of H2O that is lost from the gamet granulite textures is enough to mobilise all the Na20 and Si02. We explain the dehydration as hydroxyl being scavenged by directly adjacent migrating melts. Blattner (2005) argues that there is no need for a melt to carry the water, Na and possibly Si away from the gamet granulite textures, but rather this solute-rich fluid that is liberated in the transition to gamet granulite diffuses through the pre-existing two-pyroxene homblende granulite fabric, across a few centimetres, toward a central fracture of the texture. Export along the central surface is then rapid compared to the initial diffusion. A sink for these products is inferred to include sites of partial melting at an unknown distance, where low aH2o melt absorbs the solute-rich dehydration fluid from the gamet granulite texture. We calculated the water-saturated solidus for these rocks at approximately T = 650

and

suggest that the solute-rich fluid envisaged by Blattner (2005) would most likely bring the rocks to solidus. Furthermore, it is unlikely that such a Na20-rich solution would escape the rock without producing Na-rich minerals as metasomatic records of its passage. Furthermore, the termination of gamet granulite textures at lithological boundaries between rocks that are mineralogically similar observed in the Pembroke example (Daczko et al, 2001; Clarke et al, 2005) begs special cases of fluid flow.

The 115 Ma age of the dyke overlaps with the youngest intrusive phases of the Westem Fiordland Orthogniess and is younger than the ~120 Ma age of metamoprhic rims on zircon in the Arthur River Complex directly to the north (Hollis et al, 2003), consistent with diachronous igneous and metamorphic activity across Fiordland. The 115 Ma age of the dyke

SGGMP - Dunedin 2007

25 is ~7 million years older than the oldest ages published for the onset of extension (Scott and Cooper, 2006) and constrains the timing of gamet granulite formation for the first time.

REFERENCES Blattner, P., 2005. Transport of low-anao dehydration products to melt sites via reaction-zone networks, Milford Sound, New Zealand. Journal of Metamorphic Geology 23: 569578. Clarke, G.L., Daczko, N.R., Klepeis, K.A., and Rushmer, T., 2005. Roles for fluid and/or melt advection in forming high-P mafic migmatites, Fiordland, New Zealand. Journal of Metamorphic Geology 23: 557-567. Daczko, N.R., Clarke, G.L, and Klepeis, K.A., 2001. The transformation of two-pyroxene hornblende granulite to gamet granulite: simultaneous melting and fracturing of the lower crust, Fiordland, New Zealand. Joumal of Metamorphic Geology 19: 547-560. Hollis, J.A., Clarke, G.L., Klepeis, K.A., Daczko, N.R., and Ireland, T.R., 2003. Geochronology and geochemistry of high-pressure granulites of the Arthur River Complex, Fiordland, New Zealand: Cretaceous magmatism and metamorphism on the palaeo-Pacific Margin. Joumal of Metamorphic Geology 21: 299-313. Scott, J.M., and Cooper, A.F., 2006. Early Cretaceous extensional exhumation of the lower cmst of a magmatic arc: Evidence from the Mount Irene Shear Zone, Fiordland, New Zealand. Tectonics 25: doi: 10.1029/2005TC001890.

SGGMP - Dunedin 2007

The metamorphic evolution of the root to a magmatic arc: coronas and shear zones in the Western Fiordland orthogneiss, Doubtful Sound, New Zealand.

M. C. De PaoIi\ G. L. Clarke^ and K. A. Klepeis^

^ School of Geosciences, University of Sydney, NSW, 2006 Australia ^ Department of Geology, University of Vermont, Burlington, VT, 05405-0122, USA

The Western Fiordland Orthogneiss (WFO) is an arc-related composite batholith comprising metadiorite and metagabbro that intruded the lower crust during the early Cretaceous (126 and 109 Ma.). The development of localised garnet granulite (high-P) mineral assemblages within the WFO (P ^ 13 kbar T ~ 780°C) immediately followed batholith emplacement. The P-T evolution of the WFO has been a topic of debate in recent years; issues such as the depth of intrusion and the post-emplacement evolution remain controversial. In this study we combine detailed petrology and calculated mineral equilibria to help shed some light on the evolution of the WFO from emplacement, through to garnet granulite formation and finally exhumation.

Corona reaction textures and discrete shear zone assemblages record changing mineral paragenesis. Five discrete metamorphic events (Ml-5) can be distinguished. Ml represents amphibolite facies hydration of igneous assemblages. M2 is synonymous with a progressive dehydration of Ml hornblende in gabbroic rock compositions and involves symplectic intergrowths of clinopyroxene, biotite and quartz after igneous and Ml minerals. M3 garnet granulite is only observed within spatially localised zones of dehydration termed Garnet Reaction Zones (GRZ). Within the GRZ's gamet-clinopyroxene-K-feldspar-quartz coronas enclose all earlier formed mineral assemblages. M4 mineral assemblages are restricted to major shear zones and represent the recrystalisation and hydration of garnet granulite assemblages to form garnet amphibolite. Late stage fracturing and hydration induce the development of greenschist facies M5 mineral assemblages replacing the earlier formed

SGGMP - Dunedin 2007

textures. P-T pseudosections calculated in the chemical system NCKFMASH using THERMOCALC are consistent with Ml assemblages having formed during post-magmatic isobaric cooling at low-P. Cooling was followed by burial of the terrain and the development of transitional granulite-gamet granulite facies M2 symplectites. Further cooling at high-P resulted in the formation of garnet granulite facies assemblages in M3 GRZ's. The generation of garnet granulite was followed by a switch from burial to exhumation and cooling with concomitant development of crustal scale M4 garnet amphibolite facies shear zones. A dynamic history is inferred for the batholith: mid crustal intrusion, followed by cooling and burial contemporary with further magmatic flux. The development of localised garnet granulite reflects burial of the WFO to P— 14 kbar, and was proceeded by a switch in the tectonic regime that resulted in cooling and exhumation of the terrane along crustal-scale extensional shear zones.

SGGMP - Dunedin 2007

Granite formation in extensional accretionary orogens: the generation of maximum I-S type granite geochemical distinction.

J. Foden, M. Hand, C. Lawley, M. Burdett

Geology and Geophysics, School of Earth and Environmental Science, University of AdelaideJohn.foden@adelaide.edu.au

The concept of I- and S-type granite was coined in SE Australia (Chappell and White, 1974), based on observations and data from Siluro-Devonian rocks in the Lachlan Foldbelt on the SE margin of Gondwanaland. Although their petrogenetic interpretations are flawed (Collins, 1998), Chappell and White (1974) were clearly correct in observing that the striking mineralogical and geochemical distinction that discriminated these granite groups reflected very different source materials. This core observation of I- to S- type contrast has nevertheless been persistently difficult to transfer to other geological terrains world-wide where granites tend not to fall into such discreet baskets (Frost et al, 2002). Although this mismatch has often been attributed to "problems with translation" the reality that the tectonic setting and hence mechanism for granite formation in the eastern Gondwanan margin of Australia and Antarctica was uniquely different from that of many other orogenic belts has not been thoroughly analysed.

This regime of granite formation is quite different from that of the classic Alpine-Himalayan orogen where large volumes of lower crust are subjected to regionally elevated temperatures over time periods approaching the thermal constant of the continental lithosphere. This leads to the expansion of crustal melting domains to include a diverse range of ortho- and paragneiss sources. Likewise, granite production in the Andean setting is also different, initiated by supra-subduction melt production in the mantle wedge that then interacts in complex ways with the quite thick crust of the upper plate.

SGGMP - Dunedin 2007

29 As highlighted by Foden (2006), Collins (2002) and Gray and Foster (1997), we contend that these (perhaps necessarily east-facing?) trailing continental margin settings that face the oldest oceanic lithosphere on the globe, that have persisted in the western Pacific since the start of the Palaeozoic are very different and distinct settings for granite formation. From its inception in the Late Proterozoic, the eastern Australian margin has had an inherently extensional character, only briefly interrupted by periods of contraction. This extensional character was maintained both when East Gondwana had a passive margin with the newly opening Pacific directly following Rodinian rifting, and also after the commencement of subduction in the Middle Cambrian (Foden et al, 2006). Following the start of subduction, regular ocean-ward slab retreat resulted in development of back-arc rift basins, only briefly interrupted by short interludes of compression (Collins, 2002; Foden et al, 2006; Gray and Foster, 1997). This setting persistently creates and destroys often small, steep-sided, extensional or transtensional basins that before the start of subduction, often extended aulacogen-like into the margin of the continental craton. In the absence of subduction these rifts decompressed the asthenosphere enough to produce basah and later (from the middle Cambrian), with subduction when water was added to the mantle wedge, this mafic melt productivity was increased. Whether due to wet-melting or pure decompression, the mafic magma advected asthenospheric heat to the sedimentary fill. This created felsic magmas whose sources are mixtures of contemporary basalt and the sediment (Foden et al, 2002; Collins, 1998). I-types are products of contaminated fractionating parental mafic magmas (AFC), while S-types are due to the escape of diatexitic magma derived from the basin's sediment with variable mafic contamination (Foden et al, 2002).

A feature of this style of continental margin is that many processes remain the same whether during the initial passive extensional phase or the subsequent subduction phase (deep basin formation, sedimentary deposition, mantle-melting and sediment -basalt interaction, granite formation and high-grade metamorphism). There are a number of Late Neoproterozoic to Early Cambrian locations where it appears basin and granite formation predated subduction (514 Ma; Foden et al, 2006) off the Australian margin. These include the -725 Ma Cape Wickham granite on King Island (Berry et al, 2005), and in central Australia, in the margin of the Arunta inlier, the Irindian Basin (Buick et al, 2005). This developed as an ultra-deep (> 15km) depocentre within the Amadeus-Georgina basins at a time (545 Ma) when the shallow-water platformal Arumbera sandstone was deposited (Maidment et al, 2005). The

SGGMP - Dunedin 2007

30 thick sedimentary fill (now the Upper Stanovis Gneiss eastern Harts Range; Maidment et al, 2005) in this deep strike-slip trough was intruded (at -5.5 kbars) by mafic sills at 520 Ma producing S-type granite of the Stanovos Igneous Suite.

Further South in the Adelaide Foldbelt and extending east to the Glenelg area of western Victoria, the Kamantoo basin was another very deep Early Cambrian (520-515 Ma) strikeslip trough at the craton margin. Again I- and S-type granite formation occurred as a resuh of interaction between contemporary basin sedimentary fill and basahic mantle-derived magmas (Foden et al, 2002, 2006). I-types are the resuh of AFC processes while S-types are largescale melts of the Kanmantoo Group. This basin just predated the onset of subduction beneath the Australian margin (Foden et al, 2006) but was inverted by the Delamerian Orogeny that marked the start of subduction in this part of the western Pacific.

With the probable eastward retreat of the Delamerian subduction zone, the locus of back-arc basin formation, periodic deformation and granite formation moved to Victoria and NSW (Gray, 1997). Granites spanned a period from the Ordovician to the Carboniferous. The granites form in roughly defined belts that we would interpret to reflect a combination of mantle melt production and sediment-filled extensional basins. The relative proportion of Ito S-type is a mark of the mantle productivity.

As observed by Collins (2002), the

significant increase in granite production in the Lachlan compared with the purely extensional pre-subduction regime before 514 Ma, reflects the enhanced efficiency of mantle melting even in the back arc when subducted water is available. The distinctive tectonic setting for granite formation presented by extensional accretionary orogens is defined by the geochemistry of the granites themselves, with much more bimodal ASI frequency signatures than granite signatures in other settings.

REFERENCES Berry, R.F, Holm, O., and Steel, D., 2005. Chemical U-Th-Pb monazite dating and the Proterozoic history of King Island, southeast Australia. Australian Journal of Earth Sciences 52:461-471. Buick I.S., Hand, M., Williams, I.S., Mawby, J., Miller, J., and Nicoll, S., 2005. Detrital

SGCadP - Dunedin 2007

31 zircon provenance constraints on the evolution of the Harts Range Metamorphic Complex (central Australia): links to the Centralian Superbasin. Journal of the Geological Society of London 162: 777-787. Chappell, B.W., and White, A.J.R., 1974. Two contrasting granite types. Pacific Geology 8: 173-174. Collins, W.J., 2002. Nature of extensional accretionary orogens. Tectonics 21 (4): 6-1 to 612. Collins, W.J., 1998. An evaluation of petrogenetic models for Lachlan Fold Belt granitoids: Implications for crustal architecture and tectonic models. Australian Journal of Earth Sciences 45: 483-500. Foden, J., Elburg, M.A., Turner, S.P., Sandiford, M., O'Callaghan, J., and Mitchell, S., 2002. Granite production in the Delamerian Orogen, South Australia. Journal of the Geological Society London 159: 601-621. Foden, J., Elburg, M., Dougherty-Page, J., and Burtt, A., 2006. The timing and duration of the Delamerian Orogeny: Correlation with the Ross Orogen and implications for Gondwana assembly. Journal of Geology 114: 189-210. Frost, R., Barnes, C., Collins, W., Arculus, R., Ellis, D., and Frost, C., 2001. A geochemical classification for granitic rocks. Journal of Petrology 42: 2033-2048. Gray, D.R., 1997. Tectonics of the southeastern Australian Lachlan Fold Belt: structural and thermal aspects. In: Orogeny Through Time (Burg, J-P., and Ford, M., eds) Geological Society Special Publication 121: 149-177. Gray, D.R., and Foster, D., 1997. Orogenic concepts - Applications and definitions: Lachlan Fold Belt, eastern Australia. American Journal of Science 297: 859-891. Maidment, D., Hand, M., and Williams, L, 2005. Tectonic cycles in the Strangways Metamorphic Complex, Arunta Inlier, central Australia: geochronological evidence for exhumation and basin formation between two high-grade metamorphic events. Australian Journal of Earth Sciences 52: 205 - 215.

SGOIP - Dunedin 2007

The use of zircon in diamond exploration - a preliminary case study from the Cempaka deposit, SE Kalimantan, Indonesia

Ian T Graham^'"^, Lee Spencer^ Gregory Yaxley^ Larry Barron"^

^ School of Biological, Earth and Environmental Sciences, UNSW, Sydney NSW 2052 ^ P.T. Galuh Cempaka ^ Research School of Earth Sciences, ANU, Canberra ACT 0200 Geoscience, Australian Museum, 6 College St Sydney, NSW 2010

Introduction Borneo is one of the oldest yet least known sources of gem-quality diamonds, with workings dating as far back as possibly 600 A.D. (Spencer et al, 1988). Historic production mostly came from the westem part of Borneo using traditional hand-mining methods (Spencer et al, 1988). In 1965, a gem-quality 166.85 carat diamond, known as the 'Tri Sakti' or Three Principals' was found in SE Kalimantan, and this discovery ignited the interest of large westem mining companies.

The diamond deposits of Kalimantan occur exclusively as

paleoalluvial and present-day alluvial deposits, and despite a number of investigations (e.g. Bergman et al, 1987, 1988; Spencer et al, 1988; Taylor et al, 1990) it is still unclear as to their primary igneous source. The Cempaka diamond deposit is the only working mine and is located in SE Kalimantan, Indonesia (3^ 30'S, 114° 45'E), 12 km S of Banjarbaru, the nearest town and 40 km SE from the nearest large city of Banjarmasin.

Background The region comprises a belt of Late Jurassic to Cretaceous metasedimentary and metavolcanic rocks (Manunggal Formation), surrounded by Cenozoic sedimentary basins and overlain by Late Cenozoic fanglomerates deposited at the base of the Meratus Mountains. The Cempaka deposit is immediately contained within and derived from, erosion of the lateritised fanglomerates. Kalimantan underwent significant plate collision in the Cretaceous SGGMP - Dunedin 2007

33 and the SE highlands contain a number of sub-parallel Cretaceous ophiolite belts. Structural analysis suggests post-collision rifting. The diamonds appear to have been derived from the Meratus Mountains to the NE and reworked through several sedimentary cycles into Cenozoic and Quaternary depocentres under low-lying coastal swamps. The alluvial diamonds at the Cempaka deposit are associated with PGM, zircon, magnetite, chromite (from the ophiolites) and corundum/diaspore. No typical 'kimberlitic' indicator minerals such as pyrope garnet have been found (Spencer et al, 1988).

Although the mineral 'chromite' has been used to locate the primary igneous source in diamond exploration studies (e.g. Griffin et al, 1997), it is compromised for the Cempaka deposit as any chromite grains from the possible primary diamond source would be swamped within the deposit by the far more abundant chromite grains derived from the weathering and erosion of the nearby ophiolite belts of the Meratus Mountains. Zircon is a chemically inert and refractory phase which can survive weathering and transportation, along with the high temperature metamorphism and even some partial melting in the crust (Hinton and Upton, 1991; Corfu et al, 2003). Because of this, it has the ability to retain substantial chemical and isotopic information, long after it first crystallized (Finch and Hanchar, 2003) and so could be used in geochemical exploration.

Method Zircons were sampled from heavy mineral concentrates at the processing plant. Two distinct sample sizes were immediately apparent, a + 1mm suite and a sub-mm suite. Individual grains were then hand-picked for later scanning electron microscopy (zircon morphology), fission track dating, cathodoluminescence imaging (CL) and LA-ICP-MS.

Zircon Morphology The + 1mm grains occur as subhedral to euhedral, equant to subprismatic grains that are generally well-rounded and exhibit moderate to strong crystallographically- controlled magmatic corrosion, suggesting a xenocrystic origin. Many have well-developed percussion marks due to alluvial transportation. Most of the sub-mm zircon grains are euhedral, highly angular in shape and lack rounding, suggesting derivation from a local source.

SGGMP - Dunedin 2007

34 Zircon Fission Track Dating A sample of ten + Imm zircon grains gave a well-defined single pooled age of 132.1 ± 8.1 Ma (Early Cretaceous) while eight sub-mm zircon grains gave a well-defined single pooled age of 82.9 ±3.6 Ma (Late Cretaceous).

Cathodoluminescence Imaging Both the + Inun and sub-nun zircons generally exhibit well-developed oscillatory growth zones though some exhibit more complex sector zonation. There is no evidence of any earlier inherited cores or later metamorphic overgrowths.

Zircon LA-ICPMS analyses Results of the LA-ICP-MS analyses are summarized in Table 1. Comparison with analyses from a range of various igneous rocktypes from Belousova et al. (2002) shows that both zircon suites, and in particular the sub-mm zircon suite are most like those from lamproites.

Discussion The + 1mm zircon suite contains some moderately rounded grains with well-developed percussion marks suggesting moderately prolonged alluvial transportation (i.e. distal source). In contrast, the sub-mm grains are sharp, angular and retain fine magmatic resorption features suggesting derivation from a proximal source. The zircon FT dating clearly shows that both suites are of different age, with the + 1mm zircons being Early Cretaceous and the sub-mm zircons being Late Cretaceous in age. These suggest that magmatic activity occurred on Kalimantan in Early and Late Cretaceous times, and that breccias or eruption centres having these ages should be targeted in further exploration. Cathodoluminescence imaging show that the zircons are clearly of magmatic origin, lacking earlier inherited cores or later metamorphic overgrowths. With regards to chemistry, although there are some differences between the two suites, the maximum values for the + 1mm zircons overlap the minimum values for the sub-mm zircons. In general, the zircons can be characterized as moderate total REE, low U and low Y zircons. Based on a comparison with zircons from a wide variety of igneous sources (Belousova et al, 2002), the Cempaka deposit zircons are most akin to those from lamproites. This small study has shown that in places where no typical 'kimberlhic'

SGGMP - Dunedin 2007

indicator minerals exist, the mineral zircon can be used effectively in diamond exploration, especially when combining the techniques of electron microscopy, cathodoluminescence imaging, FT dating and LA-ICP-MS analysis. The next stage in exploration will be to take heavy mineral concentrates upstream of the Cempaka deposit and analyse zircons from each of these using the methodology undertaken in this preliminary study. It will also be beneficial to conduct U-Pb SHRIMP dating of the zircons dated using FT methods in order to determine the zircon crystallization ages and compare these to the FT ages. Table 1. Comparative geochemistry of Cempaka, lamproitic and kimberlitic zircons Type

Cempaka +lmm

Cempaka sub-mm

Kimberlite

Lamproite

No Analyses

284

0.5-1.4

4-41

0.2-6.7

2.3-30

0.1-0.5

0.7-8

0-4.2

0.5-9.5

0.1-0.4

0.2-2

0-3.4

0-3.3

1.1-4

4-41

0-10

3.3-33

7-30

22-167

0-25

7-106

3-14

9-59

0.1-7.7

1.5-36

19-87

46-274

0.4-25

5-154

60-305

112-551

0.2-36

7-277

13-72

21-96

0.1-7

1-56

0.1-0.4

0-1

0-144

0-2.1

113-496

330-1473

4-194

53-1081

0.15-0.7

0.4-4

0-23

0.16-5.2

0.6-0.8

0.7-1.2

0.6-2.3

0.8-1.3

0.04-0.3

0.2-2

0.1-13

0.07-2.1

1-48

16-300

0.8-48

12-188

7-90

32-440

3-69

17-348

Kimberlite and lamproite data from Belousova et al. (2002)

SGGMP - Dunedin 2007

Gold-variscite mineralization at Woodlands, Western Australia

Elena Hancock, Ernest Nickel^ Michael Verrall\ David Vaughan

^ CSIRO Exploration and Mining, ARRC

Location and geology The gold-variscite mineralization is located on Woodlands Station about 50 km NEE of the homestead (24°48'S, 118%6'E) in Proterozoic rocks of the Capricorn Orogen. Outcrops of the variscite were found in an anticlinal structure of the silicified shale-arenite formation of the Bangemall Subgroup. Other lithologies include dolomite and chert. Dolerite sills are developed in the area (Fig. 1).

Bangemall Siiperciroup Collier Group & conglomerate, Edmund Group Sandstone, slltstone. mudstone. dolomite, and DIsSove'ry ChS?T: ma&sfve or laminated chert. mudstone, and slltstone Mudstone, slltstone. chert, dolomite, sandstone OocalV dolomitic). and subordinate conglomerate §

Gascoyne Complex Granite, monzogranite, gniess. and amphibollte Quartz-muscwite schist muscwite schist, and phyllite Mne

Fault

Prospect

Geological boundary Homestead Locality ==

T T-tTTT-.TTTTTTy-,

L 1 i

irf*" •

' ^ ^^

EGERTON

Figure 1. Generalized geological map of the Mount Egerton 1:250 000 sheet (Morris, 2005 after Cooper e t a l . , 1998)

SGGMP - D u n e d i n

2007

Mineral occurrei

Gs, ,Au. Mn

Gemslone, Gok

50 km

Mining centre

Woodlands

•

gold-variscite

37 Mineralisation The variscite occurs in a number of conformable sub-horizontal narrow veins, less than 70 mm wide, in a 0.5 m wide zone that can be traced for at least 1 km. The veins occur in a brecciated shale and consist of massive and spindle-shaped variscite, secondary phosphate minerals, alunite, jarosite, kaolinite and hydrated iron oxides.

Variscite The variscite is translucent, with a unique dark green colour (emerald-green), and is of a high gemological quality. X-ray powder diffraction analysis of a number of specimens shows that the variscite consistently contains some admixed metavariscite, especially in the lighter cloudy green zones. For the sake of brevity, both dimorphs are simply referred to here as variscite.

Cr 52 Variation

Figure 2. "Honeycomb" or "spiderweb" texture of Woodlands variscite. Field width 30 mm.

Figure 3. Distribution of chromixam along the line AB. (measured by LA-ICP-MS)

The "honeycomb" or "spiderweb" texture seen in cross-section (Fig. 2)

is created by

irregular spindles of dark green variscite, 5-30 mm long, some of which show concentric colour zoning - sometimes with a darker, and sometimes a lighter core, and generally with a lighter-coloured rim. Individual spindles are separated by intergranular material. Laser ablation (LA-ICP-MS) analysis (by Dr.J.Watling, UWA) along a line perpendicular to the colour zoning reveals a higher concentration of the Cr 52 isotope in the dark green zones

SGGMP - Dunedin 2007

38 (Fig. 3). The intergranular material shows an increase of Cr52, as well as a number of other elements. The distribution of Fe 57 and V 51 isotopes are virtually unchanged.

These results support the conclusion Lhai the green colour of variscite is due to absorption by octahedral Cr^^ (Calas et al, 2005). The chromium content of the Woodlands variscite is up to 0.5 %, whereas the Milgun bluish-green variscite (about 70 km to the southeast) contains only 0.02 % Cr. The source of the chromium is probably chromite which occurs as a dissemination of fine grains in veinlets in the sediment. In thin section, the variscite texture can be seen to vary from microcrystalline aggregates of anhedral grains (Fig. 4a) to micronodules consisting of radial crystals (Fig. 4b).

Figure 4. Variscite microstructure (Thin-section photo micrographs) (a) fine grains; field width 250 ]jim and (b) radial crystals; field width 1mm

The micro-nodules have elevated iron content due to partial replacement of A1 by Fe^^. Some small (< 20 ^m) tabular and pyramidal crystals of variscite were seen on one fracture. The most unusual characteristic of the Woodlands variscite is that it contains visible gold.

Gold Gold is distributed unevenly in the variscite specimens, with assays up to 10 g/t, and occurences range from discrete fine grains to visible concentrations in the variscite spindles (Fig. 5). There are also larger single grains in the rock matrix - up to 300 |Lim in diameter.

SGGMP - Dunedin 2007

Figure 5. Inclusions of gold flakes in variscite spindle. Field width 10 mm.

Figure 6,

Sponge gold.

Dendrite-like sponge gold in variscite fissure. Field width 1.5mm

The composition of the gold is variable, with the silver content increasing from below the limit of detection to 15 wt% Ag. In the variscite spindles, the gold occurs as visible spongelike flakes and dendrites (Fig. 6) from ten microns to 2 mm, commonly oriented normal to the colour zoning and variscite bedding. The colour is rich yellow, and Scanning Electron Microscope Energy Dispersive X-ray Spectroscopy (SEM EDS) analyses show no silver above the detection limit, and 1-2% Cu.

The variable silver content shows that the gold mineralization is complex and is represented by at least three types of gold: Colloidal, with a small amount of silver, in variscite and rock matrix; Massive, with silver, in variscite and rock matrix; Sponge, without silver, in variscite. Multi-trace-element "fingerprinting" (LA-ICP-MS, UWA) did not show appreciable differences between the various types of gold.

SGGMP - Dunedin 2007

40 Other minerals The intergranular material between the variscite spindles includes crandallite, a fine grained crandallite-alunite-jarosite mixture, fragments of shale and quartz, and relicts of pyrite, monazite, xenotime and anatase grains. Other secondary phosphates in the variscite assemblage include wardite and turquoise crystals, and montgomeryite spheres with a crust of millisite. Hydroxylapatite occurs mainly as secondary hexagonal crystals, generally as colourless acicular crystals 0.1-0.5 mm in length; however, a primary altered grain, 0.1 mm in size in a clay matrix was also observed.

Discussion Variscite is generally considered to be a secondary mineral, and the presence of primary hydroxylapatite grains in the sediment could be the clue for the phosphate source. The different types of gold associated with the variscite probably indicate different stages of deposition. The massive gold in the matrix probably represents stratabound mineralization associated with chert and pyritic sillstone/shale, as in the Egerton, Horseshoe Lights and Labouchere epigenetic gold deposits in the Capricorn Orogen (GSWA, 1990). Stream-hosted gold from the surrounding area exhibits similar physical and geochemical characteristics. Massive gold in variscite may be the result of the in situ re-precipitation of sediment-hosted gold after the formation of the variscite. Gold could be leached from the sediment and carried as a AuCU" complex by a low-temperature solution, with high salinity, high Eh and low pH (Mann, 1984): 4Au^+16Cr + 302+ 12H^

^ 4AuCl4" + 6H20

This solution might pick up aluminium from the country rock and replace the brecciated sediment (chert/shale/sillstone) containing disseminated apatite and gold.

The precipitation of colloidal variscite occurs at 800^C at low pH (Morales et al, 1992), which is consistent with conditions favouring the precipitation of gold from solution. During deposition, the colloidal variscite masses were desiccated and shrinkage cracks formed, creating a mass of spindles with intergranular fissures. The colour zoning could be the result of this formation. Later, with slightly increasing pH, variscite began to alter, and crandallite filled the available space. Other secondary phosphates are the resuh of later variscite

SGGMP - D u n e d i n

2007

41 alteration and they commonly form crystalline aggregates in solution cavities and fractures. At the time of variscite deposition, the sponge gold precipitated from solution, and formed in the interstices between the variscite grains in the directions of the radial growth and in shrinkage fissure perpendicular to the colour zoning zones and bedding. Raising the pH or reducing AuCU" with Fe"^^ causing the precipitation of gold from solution (Mann, 1984): AuCU" +

+ 6H2O

^ Au® + 3FeOOH + 4Cr +

It is also possible that colloidal gold could be precipitated from a chloride solution by phosphorus or by phosphate compounds in the Woodlands gold-variscite formation.

REFERENCES Calas, G., 2005. The origin of the green color of variscite. American Mineralogist 90: 984990. Mann, A.W., 1984. Mobility of gold and silver in lateritic weathering profiles: some observations from Western Australia. Economic Geology 79: 38-49. Morales, J.G., Clemente, R.R., Matijevic, E., 1992. The mechanism of precipitation of colloidal variscite (AIPO4.2H2O) particles. Journal of Colloid and Interface Science 151 (2). Morris, P.A., 2005. Mount Egerton 1:250 000 map sheet. Western Australia. Regolith landscape evolution across Australia. Cooperative Research Centre for Landscape Environments and Mineral Exploration (CRC LEME). September, pp. 323-327.

SGGMP - Dunedin 2007

Timescales of magma genesis and differentiation at Lopevi Volcano, Vanuatu, SW Pacific

Heather Handley\ Simon Turner\ Ian Smith^ and Robert Stewart^

^ GEMOC, Dept. of Earth and Planetary Sciences, Macquarie University, Sydney, NSW 2109, Australia (lihandley@els.mq.edu.au). ^ Department of Geology, University of Auckland, Auckland, New Zealand ^ Institute of Natural Resources, Massey University, Palmerston North, New Zealand

Summary We present the first detailed U-series study of an individual volcanic centre within the Vanuatu arc and show that magmatic differentiation occurs over relatively rapid timescales (<1500 years) at Lopevi Volcano. Assimilation of mafic arc basement during fractionation crystallisation in the crust exerts strong control on the isotopic and geochemical composition of the erupted lavas.

Introduction Lopevi is one of the most active volcanoes in the Vanuatu (New Hebrides) intra-oceanic island arc. During historical time, eruptions have occurred from both summit and flank vents of the 7 km wide volcano and the island was permanently evacuated following major eruptions in 1939 and 1960's. Whole-rock major and trace element abundances, Sr and Nd isotopic ratios and U, Th and Ra isotopic compositions have been determined for basaltic and andesitic lavas samples erupted over the last 100 years. The data are used to constrain the processes and timescales of magma genesis and evolution beneath the volcano in order to increase our understanding of intra-oceanic subduction zone geodynamics and investigate the relationship between magma supply and eruption.

SGGMP - D u n e d i n

2007

43 Geochemistry and processes of magmatic differentiation MgO contents of Lopevi lavas erupted throughout 2000-2003 cluster at around 4-5 wt%, whereas those from earlier eruptions (during the 1930's and 1960's) exhibit a wider range and extend to more primitive MgO contents (almost 9 wt%). Decreases of MgO, CaO and Fe203 and increases in Na20, Ti02 and to a lesser extent K2O with increasing Si02 are consistent with the removal of a mineral assemblage dominated by olivine and pyroxene. The downward inflection in the data that is observed in a plot of AI2O3 versus Si02 (Fig. 1) at 51 wt% Si02 suggests that plagioclase becomes and important fractionating phase in the more evolved post-2000 lavas. This is shown quantatively by least squares modelling of the major element data (Fig 1.). Fractional crystallisation is therefore an important mechanism of differentiation at Lopevi, exerting strong control on major element 21

variations.

• pre-2000 0 post-2000

Unlike volcanic rock suites of many other arc volcanoes (Davidson et al, 2007), Dy/Yb ratios at Lopevi

oc\j

18%cpx 6% ol / 3% plag /

do not systematically decrease with

increasing Si02, demonstrating that

•

amphibole cannot be an important

lower crustal depths beneath the

• •

Ir^ = 0.2 10% plag 2% cpx 0.3% ol

/• i

fractionating mineral during magmatic differentiation at mid-

volcano.

51 S i 0 2 52

Fig. 1. AI2O3 vs Si02 for Lopevi lavas distinguished by eruption age. Least squares modelling results (Ir^) are given for the vectors shown. Percentages of mineral phases removed from the initial magma are also shown.

Sr/ Sr ratios of Lopevi lavas lie between 0.703944 and 0.704086 and are characteristic of volcanoes similarly located in the central Vanuatu arc, above where the D'Entrecasteaux Ridge is being subducted and accreted (cf northern and southern sections of the Vanuatu arc (Peate et al, 1997; Raos and Crawford, 2004)). Despite the relatively homogeneous ^'Sr/^^Sr isotope compositions, Sr isotope ratios of the lavas correlate negatively with SiOa, indicating that open system processes occur during magmatic evolution. Geochemical modelling shows that assimilation of a small-degree partial melt of mafic oceanic crust (N-MORB) during SGGMP - Dunedin 2007

44 fractional crystallisation can account well for the general array of Lopevi lava data (Fig. 2). 260 O post-2000

FC(cpx+ol). r ^^

240 - •pre-2000 220

200

- -

X"90 ^ •

AFC (bulk)

CD 180 160

/ /

PC • (cpx+ ol + Plag)

2SE

AFC (partial melt)

1 \

140 J 0.70390

0.70395

0.70400

0.70405

1 1 0.70410

Fig. 2. Ba/Th-®'^Sr/®^Sr showing fractional crystallisation (FC) and combined assimilation and fractional crystallisation (AFC) models. Tick marks on curves indicate the percentage of liquid remaining. N-MORB (Sun and McDonough, 1989) is used to represent the mafic arc basement in AFC models.

Timescales constraints: U-series isotopes Uranium series isotopes provide unique insights into the timing and rates of processes fractionating U, Th and Ra within last 380 kyrs and can therefore place timescale constraints on magma generation, transfer and storage beneath volcanoes. Time information can be extracted from U-series data using an equiline diagram. Lopevi whole rock U-Th isotope data are presented in Fig. 3a. The data lie away from the equiline, showing significant excess ^^^U (25-40%).

Excess ^^^U indicates that an event fractionating U and Th took place less than 380 kyr ago (addition of slab fluid to the mantle wedge). The relatively horizontal trend of the data suggests that at t = 0 all samples had the same Th isotope ratios and different U/Th ratios.

SGGMP - Dunedin 2007

0.6

1.0

1.4

1.8

2.2

1.3 b) C J CO o s

58,000 yrs 120,000 yrs f 51.39

1.2

51.54" 49.62 ® '

52.66 52.25 • 51.48

49.95

• pre-2000 O post-2000 1.1

1.2

1.3

1.4

Fig. 3. vs. {"®U/"2Th) . a) Eguiline diagram showing Lopevi whole rock data. b) enlarged view. SiOj contents labelled for data points. Age labels represent calculated timescales for magmatic differentiation of each vertical data series.

Variation in (^^''ih/^^^Th) may therefore reflect the time taken for magmatic differentiation (see arrows Fig. 3a). Fig. 3b shows that SiOa contents generally increase with increasing (^^•^Th/^^^Th) activity ratios in Lopevi lavas; as we might expect if the vertical trends represent time (i.e. the more evolved samples have spent more time residing in the crust). Calculations utilising the Th isotope ratios produce magma crustal residence timescales of the order 104 and 105. However (^^^Th/^^^Th) correlates with SiOj and so Th isotope ratios may

SGGMP - Dunedin 2007

46 have been affected by AFC processes, and in which case, the age information will not represent true timescales of magmatic differentiation. 6

m 0^ QC 5 ho 0.5 T 1 •• 1.5 •• CO 3 2 •• CO DC

CD CM CM

1500yrs

O^s O O ^

4 •• 8 ..

secular equilibrium

SiO2

Fig. 4. vs. SiOz for Lopevi lavas, Vertical bar labelled with years indicates the time required for to return to secular equilibrium.

The presence of ^^'^Ra excesses in the samples (Fig. 4) imply that crustal residence times of magmas at Lopevi are <8000 years and signifying that Th isotope ratios have been modified by AFC. If the magnitude of disequilibria in the mafic lavas ((^^^Ra/^^^Th) = 5.5) represents that at the onset of differentiation, short timescales are involved for magmatic differentiaion at Lopevi: <500 years for the basalts and <1500 yrs for basaltic andesites (Fig. 4). Timescales obtained from the Ra isotope data are consistent with geophysical studies indicating that crustal magma chambers small/absent in the Vanuatu arc (Iyer, 1984) and also with evolution timescales predicted by numerical models for cooling and crystallisation of magma chambers (Marsh, 1989) and AFC processes (e.g. Edwards and Russell, 1998) in the order of 102 -103 years.

SGGMP - D u n e d i n 2007

Petrogenetic model

Figure 5. Schematic representation of magmatic evolution at Lopevi Volcano. Conclusions 1. Crustal residence times of magma at Lopevi is <1500 years. 2. Assimilation of mafic arc basement during fractional crystallisation limits the timescale information that can be obtained from Th isotope ratios. 3. High MgO lavas may form part of a genetically linked suite rather than a distinct magma type. 4. This study emphasises the importance of detailed geochemical studies of individual volcanic centres.

SGGMP - Dunedin 2007

48 REFERENCES Davidson, J., Turner, S., Handley, H., Macpherson, C., and Dosseto, A., 2007. Geology 35: 787-790. Edwards, B.R., and Russell, J.K., 1998. Geology 26: 1103-1106. Iyer, H.M., 1984. Philosophical Transactions of the Royal Society London A310: 473-510. Marsh, B.D., 1989. Annual Reviews in Earth and Planetary Sciences 17: 439-474. Peate, D.W., Pearce, J.A., Hawkesworth, C.J., Colley, H., Edwards, C.M.H., and Hirose, K., 1997. Journal of Petrology 38: 1331-1358. Raos, A.M., and Crawford, A.J., 2004. Journal of Volcanology and Geothermal Research 134: 35-56. Sun, S., and McDonough, W.F., 1989. Geological Society Special Publication 42: 313-345. Turner, S., Pearce, J., Hawkesworth, C., Eggins, S., and Crawford, A., 1999. Geology 27: 963-966.

SGGMP - Dunedin 2007

Correlated, in-situ analysis of U/Pb,

and sHf in zircon from

Siluro-Devonian granite in the eastern Lachlan Orogen: constraints on juvenile additions to the continental crust

Ryan Ickert, Ian Williams

Research School of Earth Sciences, AustraUan National University Ryan.Ickert@ualberta.net

Introduction The nature and origin of granite remains an outstanding scientific problem. Where does granite come from? Is granite derived entirely from partial melting of rocks with a long crustal residence time, or is there a significant juvenile (i.e. directly mantle derived) component? How do granites differentiate, and what is the relationship between granites with apparently linear correlations between elements on variation diagrams?

Even in some well-characterized granite terrains, there continues to be much debate on these questions. One such terrane is the eastern Lachlan Orogen. These granites have been the subjects of intense study for over 30 years, and have become a "textbook" locality for granite study (e.g., Winter, 2001) however there is little consensus in the literature as to the answers to most of the questions posed above. For example, Keay et al. (1997), Collins (1998), Healy et al (2004), and

Patino Douce (1999) have suggested that S-type granites contain

appreciable quantities of directly mantle-derived material, and a number of workers, in particular Kemp et al (2007) have suggested the same for I-type magmas. Other workers, in particular Chappell (1996) has suggested that they are derived entirely from the partial melting of crustal rocks. This problem has important implications for our understanding of crustal growth: if there is substantial input of juvenile magma into granitic magmas, then they may represent a hitherto unrecognised source of new crustal material, alternatively they simply represent a mechanism by which the crust differentiates.

SGGMP - Dunedin 2007

50 Analytical Approach Zircon is a highly refractory mineral that can remain both solid and closed to diffusion of U, Pb, O, and Hf during high-grade metamorphism, partial melting and in the presence of meteoric fluids. In addition, zircon is a common accessory mineral in granitic rocks and therefore preserves evidence of magmatic processes with high fidelity. The approach in this study has been to take advantage of spatially resolved zircon isotopic analysis using highresolution SIMS to measure U/Pb and S^^O, and LA-MC-ICP-MS for sHf in order to reconstruct magmatic processes and crustal evolution. Samples have been collected from Sand I- type granites and their enclaves in the Berridale, Wagga, and Kosciuszko Batholiths, where possible re-collecting from sites that have already been characterized for whole rock geochemistry as well as Sr, Nd and O isotopes.

on SHRIMP II The measurement of oxygen isotopes at a precision amenable to solving problems in igneous rocks is a recent achievement of SHRIMP II.

Major problems, including variable

fractionation of ^^O/^^O at the source slit by the earth's magnetic field, variable fractionations induced by material and topographic contrast across the mount surface, and the production of a highly fractionated, time-dependant oxygen ion beam by the electron gun have all been overcome or minimized. We currently regularly achieve internal precisions of <0.1 %o (ISE) and an external reproducibility (spot-to-spot) of 0.3-0.4%o.

Figure 1: Typical reproducibility of a secondary zircon standard (FC-1) during the course of an

hour

analytical session. Confidence limits are Iq external errors, based on the reproducibility of the primary standard, not within-run statistics. Internal uncertainties are only slightly larger than the data points. Bulk etal (2007).

SGGMP - Dunedin 2007

data from Trail

Preliminary Results We have determined the oxygen isotope composition of over 200 zircon grains from SiluroDevonian granites of the Wagga, Berridale, and Kosciusko Batholiths. All zircon analysed for oxygen were also analysed for U-Pb. Rocks included in the dataset are -433 Ma S- and -416 Ma I-type granites of varying bulk compositions, one 414 Ma gabbro and a small dataset of zircon from the 433 Ma Cooma granite, a small pluton at the metamorphic culmination of a low-P high-T metamorphic complex. Analyses of Hf isotopes in these zircon grains are in progress. The of zircon from the gabbro (-5.5%o) is indistinguishable from that of zircon in equilibrium with the mantle and provides a reference for the oxygen isotopic composition of zircon crystallized from uncontaminated, mantle-derived magma in the eastem Lachlan Orogen. Zircon from the Cooma granite provides an estimate of the expected in zircon crystallized from a magma derived entirely by partial melting of metasedimentary rocks and early measurements indicate a value of around 9.7%o. These values provide useftil reference markers when interpreting results from granites. A first order observation is that in zircon from granites previously identified as S- or Itype on petrographic, field, or bulk-chemical grounds (Chappell and White, 2001) have distinguishable 'Ô enrichments. This can be plainly seen by the bimodal nature of the probability density diagram for all analysed zircon. Zircon grains from S-type granites are fall under the high 'Ô peak at - 9.7%o and zircon from I-type granites fall on under the low 'Ô peak at roughly 7.1%o, although there is a large amount of scatter in the median 5*^0 of populations of zircon from I-type granites. S-type granites

Figure 2: Probability density diagram for 5 18O (zircon) from granites in this study

d^^O (zircon)

SGGMP - Dunedin 2007

52 Populations of zircon separated from hand-sized samples of granite behave differently in different rocks. S-type granites generally contain zircons that have

which scatter by

w^hich is beyond what can be explained by analytical uncertainties. All populations, however, seem to have identical median values near -9.7%o. Work is underway to determine whether this scatter represents variation inherited from the source, via a viscous, poorly homogenized magma or whether it is due to true open system processes like magma mixing or contamination. I-type granites typically have a main population of zircon which have identical ^^O enrichments, and some have outliers that have higher

Preliminary work

on sHf suggests that these ^^O enrichments are correlated with ^^^Hf/^^^Hf suggesting that these granites contain zircon that has crystallized from magmas with very different isotopic characteristics.

426±6 Ma 7J±QA%o72±0A%o £Hf= -2.61 \ yg^ / . ^^ / ^^ iL^Mk Jf

J^WSIkaJr

Figure 3: Typical results of coupled, in-situ analysis of U-Pb, O and Hf isotopes from this study, (a) CL image of zircon after U-Pb analysis, (b) CL image of zircon after polishing away the U-Pb sputtering pit and analysing for

of zircon after laser ablation analysis of Hf isotopes (reported data are initial ratios). There is a small, inherited core at the centre of this zircon. The granite that these zircon crystals were separated from has a whole rock

of 9.4%o and an sNdj of-2.6.

SGGMP - Dunedin 2007

53 REFERENCES Chappell, B.W., 1996. Magma mixing and the production of compositional variation within granite suites: Evidence from the granites of southeastern Australia. Joumal of Petrology 37: 449-470. Chappell, B.W., and White, A.J.R., 2001. Two contrasting granite types: 25 years later. Australian Joumal of Earth Sciences 48: 489-499. Collins, W.J., 1998. Evaluation of petrogenetic models for Lachlan Fold Belt granitoids: implications for crustal architecture and tectonic models. Australian Joumal of Earth Sciences 45: 483-500. Healy, B., Collins, W.J., and Richards, S.W., 2004. A hybrid origin for Lachlan S-type granites: the Murrumbidgee Batholith example. Lithos 78: 197-216. Keay, S., Collins, W.J., and McCulloch, M.T., 1997. A three-component Sr-Nd isotopic mixing model for granitoid genesis, Lachlan fold belt, eastem Australia. Geology 25: 307-310. Kemp, A.I.S., Hawkesworth, C.J., Foster, G.L., Paterson, B.A., Woodhead, J.D., Hergt, J.M., Gray, C.M., and Whitehouse, M.J., 2007. Magmatic and crustal differentiation history of granitic rocks from Hf-0 isotopes in zircon. Science 315: 980-983. Patino Douce, A.E., 1999, What do experiments tell us about the relative contributions of crust and mantle to the origin of granitic magmas. In: Understanding Granites: Integrating New and Classical Techniques (Castro, A., Fernandez, C., and Vigneresse, J.L., eds.). Geological Society Special Publication: 55-75. Trail, D., Mojzsis, S.J., Harrison, T.M., Schmitt, A.K., Watson, E.B., and Young, E.D., 2007. Constraints on Hadean zircon protoliths from oxygen isotopes, Ti-thermometry, and rare earth elements. Geochemistry, Geophysics and Geosystems 8: 1-22. Winter, J.D., 2001. An introduction to igneous and metamorphic petrology. Prentice-Hall, 697 p.

SGGMP - Dunedin 2007

Permian to Jurassic Gondwanan accretion in New Zealand: constraints from basalt geochemistry and detrital zircon geochronology

Dushan Jugum, Richard Norris, J. Michael Palin

Geology Department, PO Box 56, University of Otago, Dunedin dushan@geology.co.nz

Whole rock geochemistry and Laser Ablation ICP-MS U/Pb zircon dating have been used to create new models for the formation and accretion of the Dun Mountain Ophiolite Beh (DMOB) and its neighbouring terranes in New Zealand. These terranes represent a period of punctuated accretion from the Early Permian to the Mid-Jurassic on over 400 km of the Gondwanan margin. With an early Permian supra-subduction ophiohte at the structural top (DMOB), Mid Jurassic sands and muds at the base (Aspiring/Waipapa Terrane) and an inferred Triassic imbricated accretionary wedge in between (Caples Terrane).

The Dun Mountain Ophiolite had two main stages of igneous activity, an early ocean ridge assemblage and a latter supra-subduction event. The early phase has been dated by Sivell and McCulloch (2000), at 308 Ma and the latter, which comprises the bulk of the rock mass, at 280 Ma (Kimbrough et al,

1992). The first dykes of the supra-subduction phase are

plagiogranites with amphibolite inclusions. These dykes were injected along pre-existing normal faults, believed to be remnants of initial ocean floor spreading, which were remobilised in compression before the dykes fully cooled. While these plagiogranite dykes were formed by partial melting within the lower crust, later mafic dykes can be traced to the upper mantle and are more directly related to subduction.

There are no pelagic sediments on top of the Dun Mountain Beh. Instead it is overlain by breccias and conglomerates (which are) cross-cut by the last phase of igneous activity. The

SGGMP - Dunedin 2007

55 sandstone directly overlying these has a depositional age of <260 Ma from detrital zircons, leaving a 20 Ma time-gap with no sedimentation record. However, there is evidence for significant erosion during this time, with sedimentary rocks being deposited directly on ultramafics in some locations.

Plagiogranites make up a minor amount (<1%) of the total volume of rock within the Dun Mountain Beh over most of its length. However, the proportion of felsic rocks increases toward the southern end, where plagiogranites and granodiorites become dominant. New tectonic modelling of the geometry of the DMOB implies that this variation was related to distance from the subduction zone. This suggests that there once was a felsic proto arc running the whole length of the belt, which has since been eroded due to subsequent uplift along the current plate boimdary.

On the eastern side of the DMOB, structurally underneath, there is a strip of ophiolite melange up to two kilometres wide. The basaltic blocks from this melange show similar whole rock geochemistry to the stage one ocean ridge basalts of the DMOB. However they are intruded by ocean island basalts which are absent from the DMOB and are not intruded by the supra-subduction rocks which constitute the bulk of the DMOB. The sedimentary rocks within the melange show the same modal mineralogy (QFL) and detrital zircon pattern as the sediments that overly the DMOB except they are finer-grained, implying a more distal source. Soft sediment tectonic deformation is pervasive within the melange sediments and absent from those directly overlying the DMOB. This deformation may be related to the formation of the melange.

Further east of the melanges is the Caples Terrane, a package of volcaniclastic rocks with ocean floor basalts forming their structural base. Detrital zircons show that sandstones directly overlying the basal seafloor have a depositional age of less than 223 Ma. Detrital zircons from locations apparently higher in the Caples Terrane exhibit depositional ages less then 240 Ma. Due to the volcanic source for the sediment, these maximum ages are thought to be close to depositional ages as a first approximation. The basalt whole rock geochemistry from within the Caples Terrane is diverse showing ORB at the base and OIB and suprasubduction basalts in other formations.

SGC3MP -

Dunedin

2007

The Aspiring Terrane Ues structurally beneath the Caples Terrane and ranges from EarlyTriassic to Mid-Jurassic (<155 Ma) in age as determined by detrital zircons. This is within the lower greenschist facies of the Haast Schist which at that location reached the Ar retention temperature at 145-139 Ma (Gray and Foster, 2004). This gives <10-16 Ma for these zircons to form, erode, metamorphose and be cooled to the Ar retention temperature.

Figure: Detrital zircon ages across the DMOB, Caples and Aspiring terranes, and a generalised cross section across the terranes. Cross section after Mortimer (1993). Ages for the DMOB after Sivell and McCulloch (2000), and Pallai et al (1991). Caples stratigraphy after Roser and Mortimer (1993). REFERENCES Adams, C. J., Campbell, HJ., et al., 2007. Provenance comparisons of Permian to Jurassic tectonostratigraphic terranes in New Zealand: perspectives from detrital zircon age patterns. Geological Magazine: In Press. Gray, D. R., and Foster, D.A., 2004. ^^Ar/^W thermochronologic constraints on deformation, metamorphism and cooling/exhumation of a Mesozoic accretionary wedge, Otago Schist, New Zealand. Tectonophysics 385: 181-210. Kimbrough, D. L., Mattinson, J.M., et al., 1992. Uranium-lead ages from the Dun mountain ophiolite belt and Brook Street terrane. South Island, New Zealand. Geological Society of America Bulletin 104: 429-443. Mortimer, N., 1993. Geology of the Otago schist and adjacent rocks. Institute of Geological SGGMP - Dunedin 2007

57 and Nuclear Sciences geological map 7. 1: 5 000 000. Pillai, D. D. L., Landis, C.A., et al., 1991. Permian ammonoids from the Greville Formation, Upukerora Valley, Southland. New Zealand Journal of Geology and Geophysics 35: 365-372. Roser, B. P., Mortimer, N., et al., 1993. Geology and geochemistry of the Caples Terrane, Otago, New Zealand; compositional variations near a Permo-Triassic arc margin. South Pacific sedimentary basins. F. Ballance Peter. 2: 3-19. Sivell, W. J. and McCulloch, M.T., 2000. Reassessment of the origin of the Dun Mountain Ophiolite, New Zealand; Nd-isotopic and geochemical evolution of magma suites. New Zealand Journal of Geology and Geophysics 43 (2): 133-146.

SGCSMP -

Dunedin

2007

A U-Pb and hafnium in-situ zircon investigation of orthogneiss and paragneiss units of Western Fiordland, New Zealand

Luke A. Milan\ Nathan R. Daczko\ Ian Tumbull^ Andrew Allibone and Geoffrey Clarke^

^ GEMOC Key centre, Department of Earth and Planetary Sciences, Macquarie University, Sydney, 2109, Austraha ^ Institute of Geological and Nuclear Sciences, Dunedin, New Zealand ^ School of Geosciences F05, University of Sydney, Sydney, NSW 2006, Australia

Introduction Recent fieldwork in Southern Fiordland has revealed three previously unrecognised plutons within the Western Fiordland Orthogneiss (WFO) batholith (Allibone et al, 2005): (i) coarse banded gamet-clinopyroxene and plagioclase-garnet-clinopyroxene gneiss, referred to as the Breaksea Gneiss; (ii) a heavily retrogressed unit, typically homblendite and plagioclasehomblende gneiss, referred to as the Resolution North Orthogneiss; and (iii) homogenous orthogneiss, known as the Malaspina Gneiss, in which igneous textures are patchily preserved. In units (i) and (ii), igneous textures are completely destroyed by metamorphic recrystallisation during deformation. Unit (iii) is the southern continuation of the Malaspina Gneiss found and originally named in Doubtful Sound (Oliver, 1980). Dykes of the Malaspina unit (iii) cut layering within the Breaksea Gneiss.

New U-Pb age constraints The WFO shows a general younging trend towards the south. The very northern extent of the WFO batholith is oldest. For example, Mt Daniel exposes orthogneiss with 123.6 ± 3.0, 121.8 ± 1.7 and 120.0 ± 2.6 Ma ages (Hollis et al, 2004). A sample from Lake Brownlee, analysed in this study, is dated at 122.21 ± 0.5 Ma and confirms the older ages of the batholith toward the north. The batholith ages of undifferentiated WFO comparable to the Malaspina Gneiss

SGGMP - Dunedin 2007

are younger to the south with new ages determined at Southerland Sound (115.65 ± 0.7Ma), Nancy Sound (115.04 ± 0.7 Ma) Wet Jacket Arm (116.71 + 0.8 Ma), and Doubtful Sound (114.2 ±1.9 Ma). U-Pb results show that on Resolution Island, the Breaksea and Malaspina units are indistinguishable on the basis of age: 116 ± 3.1 Ma and 115.5 ± 1.2 Ma for two samples of the Breaksea unit, and 114.07 ± 0.5 for the Malaspina unit, (see Fig. 1). A plagioclasehomblende-bearing sample mapped as Resolution North Orthogneiss from Resolution Island has a zircon population with a broad spectrum of significant age peaks at ~ 2450, 2300, 2200, 960, 770, 600 and 550 Ma. A single zircon grain with a core of 118 Ma and a rim of 110 Ma was also analysed. This samples bulk rock chemistry and field relations suggest it is affiliated with the WFO.

Figure 1: Concordia plots. Top left: Malaspina Gneiss from Resolution Island, (n = 41, 114.07 ± 0.5 Ma). Top right: Breaksea gneiss from resolution Island, (n = 23, 115.5 ±1.2 Ma). Bottom left: Resolution North Orthogneiss from Coal River, (n = 27 100.1 ± 1.6 Ma). Bottom right: WFO from Southerland Sound, (n - 115.65 ± 0.7 Ma) A sample of Resolution North Orthogneiss homblendite from Coal River (north of Resolution Island) has an age of 100.1 ± 1.6 Ma (see Fig. 1). This age is significant in that it is SGOdP - Dunedin 2007

60 considered to be part of the WFO suite. Its age is considerably younger than those previously described. The hafnium isotope ratios for this sample are distinct from the rest of those analysed in the WFO suggesting this sample has not been reset as shown below.

New Hf isotope data The hafnium isotopes ratios are relatively homogeneous for all the samples analysed across the various mapped units of the WFO. They share a common model (T^ DM) source age at around -750 Ma. The Homblendite from Coal River has a juvenile component not observed in the rest of the WFO snf values, consistent with a more juvenile source from the mantle (see Fig. 2).

20.00

15.00 10,00

5.00

0.00

-GHUR - Depleted mantle ® Resolution North Orthognefss o Supper Cove Orthognefss • Malaspina and undifferentiated WFO Breaksea Gneiss

GHUR

^5.00 --10.00

-.15.00 ^20,00 50

100 Age (Ma)

150

200

Figure 2: Snf values for the bulk of the units that make up the WFO show some homogeneity, while the sample of Resolution North Orthogneiss from Coal River is distinct in age and has a juvenile component.

Hf isotopes in zircon are used here to test any potential for disturbance of U-Pb ages determined for inherited zircons in the George Sound Paragneiss (Hollis et al, 2004). The consistent hafnium model source ages (T^^dm) of approximately -720 Ma show the George

SGGMP - Dunedin 2007

61 Sound Paragneiss to be distinct from Palaeozoic paragneiss of the Tuhua Sequence, which have two distinct model source ages (T"dm) of approximately 1560 Ma and -3000 Ma, confirming the interpretation of a Mesozoic age for the George Sound Paragneiss.

REFERENCES: Allibone, A., Tumbull, L, Milan, L., Carroll, S., and Daczko, N., 2005. Relationships between granulite fades Western Fiordland Orthogneiss and older rocks in Southwest Fiordland. Program and abstracts. Geological Society of New Zealand 50th Annual Conference, Kaikoura, New Zealand. Geological Society of NZ Miscellaneous Publication 119 A. Hollis, J.A., Clarke, G.L., Klepeis, K.A., Daczko, N.R., and Ireland, T.R., 2004, The regional significance of Cretaceous magmatism and metamorphism in Fiordland, New Zealand, from U-Pb zircon geochronology. Journal of Metamorphic Geology 22: 607-627. Oliver, G. J. H., 1980. Geology of the granulite and amphibolite facies gneisses of Doubtful Sound, Fiordland, New Zealand. New Zealand Journal of Geology and Geophysics 23:27^1.

SGGMP - Dunedin 2007

Miocene-Quaternary evolution of SW Pacific arcs and backarc basins: new ages and a choice of tectonic models

Nick Mortimer\ Mike Palin^ and Rick Herzer^

^GNS Science, Private Bag 1930, Dunedin ^ Department of Geology, Otago University, PO Box 56, Dunedin ^ GNS Science, PO Box 30368, Lower Hutt E-mail: n.mortimer@gns.cri.nz

The

Taupo

Volcanic

Zone,

Kermadec-Tonga Ridges, Havre Trough and Lau Basin are the Raoul Island

volcanic arc and backarc basins caused by present day PacificAustralian plate subduction. Their extinct predecessors are arrayed to

Pacific Plate

the west.

Recently published Ar-Ar ages from the offshore New Zealand region include a 1.1 ± 0.4 Ma (2a) whole rock age on a Havre Trough basalt, 1.2 i 0.8 Ma plagioclase age on an easternmost Northland Plateau (= Colville Ridge related) basalt, 19-26 Ma ages from Norfolk and South Fiji Basin scarps and DSDP holes, 15-21 Ma ages from sodic and potassic seamounts in the Norfolk and South Fiji Basins, and 15-32 Ma ages from arc lavas dredged from the Three Kings Ridge and Northland Plateau (Mortimer et al, 2007).

SGGMP - Dunedin 2007

63 Plutonic xenoliths have been long been recognised from the 3.7 ka Matatirohia Tephra of Raoul Island (Thomas, 1888; Lloyd and Nathan, 1981; Worthington, 1998). We report a new LA-ICP-MS U-Pb age of 1.25 ± 0.06 Ma (2a, MSWD= 3.3, n=35/35 grains) for zircons from a granitoid xenolith. This is far older than the age of the enclosing tuff. Clearly the granitoid does not represent Gondwana or Paleogene arc basement (the hypothesis we originally set out to test), but it does show that the recently erupted dacites are not the first cycle of intraoceanic silicic magmatism in the Kermadec Islands.

The three Quaternary ages indicate that volcanism was active at c. 1.2 Ma across the full width of a then-actively-rifting Colville-Havre-Kermadec system. Magmatism on the Colville Ridge ranges back to 5.5 ± 0.2 Ma (based on one other whole rock K-Ar age; Adams et al, 1994). In onland North Island, there is a fairly continuous record of Northland-TVZ arc magmatism from 23-0 Ma but, except in the Lau Islands, there is no record of offshore arc (or intraplate) volcanism in the interval 14-5 Ma. And in the offshore basins there is no record of backarc basin spreading in the interval 19-1 Ma. Either the missing Late Miocene arc and backarc rocks are yet to be dredged or drilled on the Colville Ridge and in the eastern South Fiji Basin. Or for some as yet unexplained reason, arc volcanism and/or back arc spreading in the SW Pacific shut down for a period of between 9 and 18 million years while PacificAustralian plate convergence continued.

REFERENCES Adams, C.J., Graham, I.J., Seward, D., and Skinner, D.N.B., 1994. Geochronological and geochemical evolution of late Cenozoic volcanism in the Coromandel Peninsula, New Zealand. New Zealand Journal of Geology and Geophysics 37: 359-379. Lloyd, E.F., and Nathan, S., 1981. Geology and tephrochronology of Raoul Island, Kermadec Group, New Zealand. New Zealand Geological Survey bulletin 95: 105 p. Mortimer, N., Herzer, R.H., Gans, P.B., Laporte-Magoni, C., Calvert, A.T., and Bosch, D., 2007. Oligocene-Miocene tectonic evolution of the South Fiji Basin and Northland Plateau, SW Pacific Ocean: evidence from petrology and dating of dredged rocks. Marine Geology 237: 1-24.

SGOdP - Dunedin 2007

64 Thomas, A.P.W., 1888. Notes on the rocks of the Kermadec Islands. Transactions of the New Zealand Institute 20:311-315. Worthington, T., 1998. Geology and petrology of Raoul Volcano: magma genesis and fractionation processes beneath the Tonga-Kermadec arc. Unpublished PhD thesis, University of Auckland.

SGGMP - Dunedin 2007

Low-temperature metamorphic monazite: occurrence, chemistry and geochronology

Janet R. Muhling\ Birger Rasmussen^ and Ian R. Fletcher^

^ Centre for Microscopy, Characterisation and Analysis, University of Western Australia ^ School of Earth and Geographical Sciences, University of Western Australia.

Introduction Recent reviews of monazite occurrence and geochronology have concentrated on igneous monazite, and metamorphic monazite from amphibolite facies and higher grades. According to these reviews, monazite is 'relatively rare in low-grade pelitic metamorphic rocks' (Williams et al, 2007). Perhaps because of this view, there have been few geochronological studies of low-temperature monazite. For example, Harrison et al (2002) found that 'the scarcity of the mineral in most low-grade metamorphic terranes preclude (sic) widespread application' [of U-Th-Pb geochronology]. In fact, monazite is relatively common in very low grade and low-grade (T <400°C, prehnite-pumpellyite to lower greenschist facies) metasedimentary rocks, and also in hydrothermal ore deposits. In Westem Australia, we have found low-grade metamorphic monazite in metasedimentary packages throughout much of the state (Fig. 1), and have used it to give precise and accurate ages for metamorphism and, in some cases, deformation in these sequences.

SGGMP - Dunedin 2007

1. Stirling Range Fm 2. Mount Barren Group 3. Ravensthorpe Greenstone Belt 4. Yelma Fm, Earaheedy Group 5. Yelma Fm, Earaheedy Group 6. Maraloou Fm, Mooloogool Group 7. Jack Hills Greenstone Belt 8. Edmund Group 9. Mount McRae Shale, Hamersley Group 10. Jeerinah Fm, Fortescue Group 11. Mount McRae Shale, Hamersley Group 12. Jeerinah Fm, Fortescue Group 13. Jeerinah Fm, Fortescue Group 14. Hardey Fm, Fortescue Group 15. Jeerinah Fm, Fortescue Group 16. Soansville Group 17. Mount Roe Basalt, Fortescue Group

Figure 1. Locations of low-grade metamorphic monazite dated by in situ U-Pb SHRIMP analysis in Western Australia. Occurrence The form of low-temperature metamorphic monazite varies according to the grainsize of its host sedimentary rock. In shales, monazite typically forms black or grey porphyroblasts up to several hundred microns across, but millimetric nodules have been described from placers in Wales, Brittany and Belgium (e.g. Donnot et al, 1973; Read et al, 1987; Bumotte et al, 1989). The porphyroblasts are riddled with inclusions of the fine-grained micas that make up the matrix of the metapelites, and also commonly have inclusions of quartz, pyrite, zircon, Ti and/or Fe oxides, and organic matter. The porphyroblasts may be equant and irregular (Fig. 2A), or elongate and aligned with the parting in the shale (Fig. 2B), and in most cases, the monazite grew after compaction of the sediment (Rasmussen et al, 2001). The micas in the shales, and within the monazite, are generally fine-grained muscovite and chlorite, rather than clay minerals, indicating that monazite grew as a result of very low or low-grade metamorphism, not diagenesis. In metasandstones, monazite is more abundant in samples with a micaceous matrix than in orthoquartzites. In these samples, the monazite forms grains that wrap around quartz and other clasts, filling pores originally occupied by clays (Figs. 2C and 2D). Metamorphic monazite from both fine- and coarse-grained sedimentary rocks may have cores of detrital grains (Figs. 2E and 2F). The detrital cores are generally inclusion free and can be detected with high-contrast backscattered electron (BSE) images and x-ray element maps. Where detrital cores have been partially resorbed and surrounded by SGGMP - Dunedin 2007

67 metamorphic monazite, there are commonly inclusions of very fme-grained ThSi04 (huttonite or thorite) at the boundary between the detrital core and metamorphic rim (Rasmussen and Muhling, 2007). Metamorphic monazite may show a textural relationship with deformational fabrics, e.g. it may be aligned with a cleavage (Figs. 2G and 2H) or may overgrow the minerals that define a foliation. Dating of this monazite can place constraints on the timing and duration of orogenic events (e.g. Rasmussen et al, 2002).

Figure 2. A. Equant, inclusion-filled monazite. B. Elongate monazite aligned with shale fissility. C. Monazite growing around quartz and K-feldspar clasts in sandstone. D. Monazite with inclusions of quartz, zircon and pyrite in conglomerate. E-F. Monazite with solid (?detrital) cores and inclusion-filled rims in siltstone. G-H. Monazite at a sandstone-siltstone boundary with crystal faces parallel to slaty cleavage (Sc).

Chemistry The chemistry of low-temperature metamorphic monazite is distinctly different from that of its higher temperature counterparts. Low-temperature metamorphic monazites generally have low Th02 contents (<2 wt%) although the distribution is patchy and commonly shows higher values on the margins of grains (Figs. 3A and 3B). In some cases, patches of high Th may

SGGMP - D u n e d i n 2007

68 indicate the presence of detrital grains that have been resorbed and incorporated into the metamorphic monazite, accompanied by complete isotopic resetting (Fig. 3B). Metamorphic monazites, both low and higher temperature, have significantly more EU2O3 than igneous monazite, with generally slight negative Eu anomalies (Eu/Eu* 0.4-0.6). Low-temperature monazites show more within-grain variation in REE than higher temperature grains, and many grains show pronounced zoning of REE (e.g. Read et al, 1987; Rasmussen et al, 2007). In shales, some grains have cores with very low La and complementary high Nd and other MREE, including Eu (up to ~ 1.5 wt% EU2O3), while rims have high La and low MREE. The cores and rims may have crystallised at different times (e.g. Figs. 3A and 3B) or have indistinguishable ages. A similar effect is seen in metamorphic monazite that has grown around a detrital core: a ring of La-poor, MREE-vich monazite forms around the detrital grain, and is succeeded by a La-rich and MREE-poor rim (Fig. 4). The REE zoning in lowtemperature monazite results in two distinct chondrite-normalised distribution patterns (Fig. 4): the La-poor analyses have convex-upwards patterns peaking at Sm or Nd, while the Larich rims show a pattern of LREE enrichment and monotonic decrease to Eu, and flatter HREE (Gd-Dy). The HREE are less abundant than in monazites from igneous or amphibolitefacies grains. Low-temperature monazites from metasandstones have a similar range in composition to those from shales.

Figure 3. A. BSE image showing locations of SHRIMP analysis spots, and x-ray element maps for La, Sm and Th in a strongly zoned, elongate monazite crystal. B. BSE image and element maps as for A for an equant but irregular monazite grain. Note the patchy high Th values in the core.

SGGMP - Dunedin 2007

69 Geochronology Monazite is an excellent geochronometer for low-temperature events. It is resistant to isotopic resetting and does not incorporate common Pb during growth. It also has the ability to record multiple events through complex metamorphic and orogenic histories. Despite the abundant inclusions within low-temperature metamorphic monazite, particularly in shales, clear patches large enough for geochronology by SHRIMP (or EMPA) can generally be found. It is recommended that these areas be mapped for elements such as La, Sm, Th and Y to identify compositional zones that may have different ages. U-Th-Pb dating of these zones can then provide accurate and precise dates for low-grade metamorphism, and associated orogenic events. 1000000

100000

1 —»

10000

Relict core Low-La ring ^ High-La rim 1000

ii;

i i

Figure 4. Chondrite-normalised (CI chondrite values from McDonough and Sun 1995) REE pattems and x-ray element maps of a composite monazite grain with a detrital core surrounded by a ring of low-La metamorphic monazite and a broader rim of high-La monazite.

REFERENCES Bumotte, E., Pirard, E., and Michel, G., 1989. Genesis of gray monazites: Evidence from the Paleozoic of Belgium. Economic Geology 84:1417-1429. Donnot, M., Guigues, J., Lulzac, Y., Magnien, A., Parfenoff, A., and Picot, P., 1973. Un

SGGMP - Dunedin 2007

70 nouveau type de gisement d'europium: la monazite grise a europium en nodules dans les schistes paleozoi'ques de Bretagne. Mineralium Deposita 8:7-18. Harrison, T.M., Catlos, E.J., and Montel, J-M., 2002. U-Th-Pb dating of phosphate minerals. Reviews in Mineralogy and Geochemistry 48: 523-558. McDonough, W.F., and Sun, S-s, 1995. The composition of the Earth. Chemical Geology 120:223-253. Rasmussen, B., and Muhling, J.R., 2007. Monazite begets monazite: evidence for dissolution of detrital monazite and reprecipitation of syntectonic monazite during low-grade regional metamorphism. Contributions to Mineralogy and Petrology {in press). Rasmussen, B., Bengtson, S., Fletcher, I.R., and McNaughton, N.J., 2002. Discoidal impressions and trace-like fossils more than 1200 million years old. Science 296:1112-1115. Rasmussen, B., Fletcher, I.R., and McNaughton, N.J., 2001. Dating low-grade metamorphic events by SHRIMP U-Pb analysis of monazite in shales. Geology 29:963-966. Rasmussen, B., Fletcher, I.R., and Muhling, J.R., 2007. In-situ U-Pb dating and element mapping of three generations of monazite: Unravelling cryptic tectonothermal events in low-grade terranes. Geochimica et Cosmochimica Acta 71:670-690. Read, D., Cooper, D.C., and McArthur, J.M., 1987. The composition and distribution of nodular monazite in the Lower Palaeozoic rocks of Great Britain. Mineralogical Magazine 51:271-280. Williams, M.L., Jercinovic, M.J., and Hetherington, C.J., 2007. Microprobe monazite geochronology: Understanding geologic processes by integrating composition and chronology. Annual Reviews in Earth and Planetary Sciences 35: 137-175.

SGGMP - Dunedin 2007

Young trachyandesite lava domes of the Gisborne area, central Victoria, Australia: products of shallow or deeper crustal contamination of tholeiitic basaltic magmas?

Ian Nicholls, Adam Moir and Zarah Heyworth

School of Geosciences, Monash University

Location and nature of occurrences The Mt. Gisbome volcanic complex, 60 km northwest of Melbourne, is believed to contain the most diverse range of lava compositions of any eruptive centre within the Late CenozoicRecent dominantly basaltic lava fields of central and westem Victoria (Edwards and Crawford, 1940). This is a low shield-like lava cone >4 km in diameter but only a few hundred metres in height. Basaltic lavas range from highly mafic and alkaline (Ne-hawaiite) through mildly alkaline (hawaiite) to strongly tholeiitic (01- and Qz-tholeiites). The later lavas of the complex (forming the Mt. Gisbome summit and McGeorge Hill and within the lava sequence at the "Giants Grave" locality on Jackson's Creek,

km northeast of Mt.

Gisbome) are much more silicic and have been classified as either basaltic icelanditesicelandites or (according to the alkalis-silica classification scheme - Fig. 1) trachyandesitestrachydacites.

Mt. Gisbome is flanked to the north and east by a group of much smaller steep-sided lava cones, typically <1 km in diameter and <100 m high, resting on more extensive lavaagglutinate aprons (e.g. Mt. Aitken, Red Rock, Fitzgerald Hill). These are apparently monogenetic cones, formed from a single magma composition within the range 01- to Qztholeiite to icelandite/trachyandesite, the latter types similar to those of the Mt. Gisbome complex. To the northwest of Mt. Gisbome is the large Mt. Bullengarook hawaiite lavascoria cone complex, source of a narrow ridge of lava flows which extend -30 km southward to the Bacchus Marsh area. A Mt. Bullengarook lava dated at -3.5 Ma (McKenzie et al, 1984) provides the only age constraint on volcanic activity in the area, but it is not known

SGGMP - Dunedin 2007

72 whether the Gisbome group of centres is of similar age.

The Gisbome-area volcanoes he within an area of uphft just outside the northwestern margin of the Port Philip depression and the bounding Rowsley Fault. They are - 1 0 km south of the Late Devonian Mt. Macedon Igneous Complex, with its remnants of a large peraluminous gamet-cordierite-hypersthene dacitic ignimbrite pile, intruded by small granodiorite-granite plutons.

Petrographic features The trachyandesites and trachydacites (-55-65% Si02 - Fig. 1) of Mt. Gisbome itself provide clues to the genetic relationships between these lavas and the basaltic spectrum in the area. Most contain cm-scale fine grained enclaves with "quench" textures dominated by plagioclase, orthopyroxene, less abundant Ca-clinopyroxene and minor olivine. These have bulk compositions closely similar to those of the Qz-tholeiite lavas of the monogenetic cones (e.g. Deverall Hill), indicating that the parent magmas for the more silicic lavas were of similar composition and that magma mingling was involved in magmatic evolution and eruption.

< Nepheline Hawaiite Hawaiites

• Trachyandesite Suite

Trachydacite Matrix

^ -^Phonotephrite

ofN o

Rhyolite

Nephrite C (oi<io%;

iBasanite

|(01>10% Basalt

Basaltic andesite

55 60 Si02 (wt%)

Fig. 1: Classification of Gisbome-area lavas

SGGMP - Dunedin 2007

73 In hand sample, it is immediately clear that the more silicic lavas contain high proportions (to -20 vol. %) of coarse (to

mm) feldspar and quartz crystals, in addition to the common

small enclaves. In thin section, the feldspars are seen to be typically extensively resorbed, often forming highly embayed and sieved "brain"-like shapes (Edwards and Crawford, 1940) with broad melted rims charged with tiny glass inclusions.

In some cases, both sodic

plagioclase (--An25) and K-rich alkali feldspar crystals are present, which along with quartz crystals suggest a granitic or felsic volcanic source. The lavas also contain obvious xenoliths of quartz-rich Ordovician sedimentary country rock plus an enigmatic group of small plagioclase-rich xenoliths with green spinel and in rare cases garnet. The nature and source of these latter xenoliths is of particular interest, since they could possibly represent fragments of "S-type" felsic igneous material related to the nearby Mt. Macedon Igneous Complex (as could also be the case for the abundant xenocrystic feldspar and quartz), or alternatively granulitic rocks from deeper in the crust.

A further enigmatic feature of the icelandites/trachyandesites is the presence of abundant Ferich (~Mg5o) orthopyroxene crystals, often extensively embayed and sieved by melting and/or forming the cores of reversely zoned grains with more magnesian rims (-Mgjs). It is unclear whether these are xenocrysts derived from crustal rocks or crystals formed in "hybrids" reflecting mingling/mixing between more and less evolved magmas of the Gisbome suite.

Trace element and isotope geochemistry Tholeiitic basalts of the Gisbome suite are geochemically akin to other Late Cenozoic-Recent tholeiitic basalts of SE Australia, with the clear oceanic island basalt (OIB) trace element and isotope affinities of these (Fig. 2). An analysed Qz-tholeiite enclave from the trachyandesites has a trace element pattern almost identical to those of the Qz-tholeiite lavas.

SGGMP - Dunedin 2007

100

Tholeiitic basalts

Nb Ta

Nd Sm Zr

Gd Tb

Er Tm Yb

Fig. 2: Incompatible trace element patterns of Gisbome suite Qz-tholeiites and an enclave from Mt. Gisbome trachyandesites.

Within the trachyandesite group, levels of "crustal" elements such as K, U, Th and Pb are strongly correlated with Si02 content, and with increasing silica, positive anomalies

100

Fig. 3:

Incompatible trace element pattems of Gisbome suite trachyandesites, showing

pronounced Pb and Nb/Ta anomalies.

associated with Pb and negative anomalies with Nb and Ta become increasingly pronounced (Fig. 3).

Sr and Pb isotopes also show strong positive correlations with Si02 (Fig. 4),

supporting the suggestion of progressive "contamination" by felsic crustal material.

SGGMP - Dunedin 2007

0.7040

0.7045

0.7050

0.7055

''Srl^'Sr

0.7060

0.7065

Fig. 4: Correlation between Si02 content and ^^Sr/^^Sr ratio for Gisbome suite Qz-tholeiite to trachydacite lavas. Conclusions The lavas of volcanic centres within the Gisbome area are highly diverse, ranging from Nehawaiite through hawaiite, 01-tholeiite and Qz-tholeiite to unusual more silicic types (icelandites or trachyandesites). The silicic lavas contain a prominent component of xenocrystic Na-plagioclase, K feldspar and quartz, probable xenocrystic orthopyroxene and quartz-rich sedimentary and plagioclase-rich, spinel- and garnet-bearing metamorphic or igneous xenoliths. The progressive addition of these components has had a pronounced effect on major and trace element and isotope geochemistry, with levels of "crustal" elements, plus Sr, Nd and Pb isotopes, strongly positively correlated with Si02 content. The origins of the xenocrystic and xenolithic components, with the exception of obvious low grade quartz-rich sedimentary xenoliths, are at this stage enigmatic. They could have been derived from unexposed portions of the Late Devonian Mt. Macedon Igneous Complex dacitic ignimbrites and granites, or possibly from felsic mid-crustal granulitic rocks (with possible implications for the basement of the Lachlan Fold Belt).

SGGMP - Dunedin 2007

76 REFERENCES Edwards, A.B., and Crawford, W., 1940. The Cainozoic volcanic rocks of the Gisbome district, Victoria. Proceedings of the Royal Society of Victoria 52: 281-311. McKenzie, D.A., Nott, R.J., and Bolger, P.P., 1984. Radiometric age determinations. Reports of the Geological Survey of Victoria 74: 1-65.

SGGMP - Dunedin 2007

Origin of compositional variations in TTG-like magmas: an integrated experimental and numerical study

T. Rushmer \ A. Getsinger^, M. D. Jackson^

^ GEMOC, Department of Earth and Planetary Sciences, Macquarie University, 2109 NSW Sydney Australia Department of Geology, University of Vermont, USA ^ Department of Earth Science and Engineering, Imperial College, UK

Major, trace, and REE compositions of both Archean TTGs and modem adakite-like magmas have been used in conjunction with batch melting experiments and models to infer source rock compositions, depths of melting, and tectonic setting (e.g. Martin and Moyen, 2002; Condie, 2005; Rollinson, 2006). However, the impact on the magma geochemistry during physical melt segregation processes has not been considered to date. When melt initially migrates, it can interact with its partially molten host and this process may have a profound impact on the composition of the segregated melt as it leaves the source region.

In this study, we are investigating through numerical modelling and experimental testing, the hypothesis that TTG arc crust formation is not only a function of partial melting of a mafic source region, but the time and length scales for melting and extraction, and melt segregation mechanisms themselves. In the experimental investigation we have designed melt segregation equilibrium (MSE) experiments to reproduce the local changes in bulk composition that are predicted to occur in response to buoyancy-driven melt segregation along grain edges and associated compaction of the solid residue.

Numerical modeling In the continental crust, numerical models suggest that the steep thermal gradient through the source region will dominate the compositional evolution of melt and matrix; pressure

SGGMP - Dunedin 2007

78 variations are small because of the restricted vertical extent over which melting occurs (Jackson et al, 2005). If melt segregates via fractures, which efficiently drain melt before it has interacted with the matrix, then segregated meh compositions may approach the limit of fractional melting. However, if melt flows along grain boundaries, then it will migrate through a temperature field which varies spatially and temporally, whilst interacting thermally and chemically with the surrounding matrix. Jackson et al (2005) considered partial melting caused by the intrusion of hot, mantle derived basaltic magma into basaltic lower crust. The model represents one end-member of a spectrum, at the other ends of which are models where melt segregation is controlled entirely by deformation or fractures. At time zero, the intrusion of magma causes partial melting of the overlying rock. This zone of partial melting is the source region; its top is defined by the position of the solidus isotherm and its base by the contact with the underlying magma. As the melt migrates upwards along grain boundaries, it thermodynamically equilibrates with compacting matrix migrating downwards and the composition of the melt continually evolves as it migrates upwards. The numerical model predicts a combination of melt migration and chemical reaction through a steep thermal gradient. Hot meh migrating upwards thermodynamically equilibrates with cool matrix migrating downwards, so the compositions of both continually evolve: higher temperature components freeze out of the melt, while lower temperature components melt out of the matrix.

This yields meh compositions at the top of the source region (where

temperatures are low) which correspond to small degrees of equilibrium partial melting (f < 5 - 15 vol%) of the source rock, but which have accumulated in large volumes in a zone of high porosity (f >40 vol%).

Experimental Study To test the model, we begin by conducting a normal equilibrium (batch) partial melting experiment on a natural basaltic bulk starting composition, at pressures appropriate for deep crustal melting (>1.0 GPa). We term this a 'direct partial melting' (DPM) experiment. We choose a temperature that will yield the same degree of equilibrium melting (f) as that predicted in the high porosity zone by the numerical model. We then determine the solid phase mineralogy and partial melt composition for this degree of melting.

We make a

synthetic glass that represents this partial melt composition. We now have the starting materials for the next partial melting experiment, which we term a 'meh segregation equilibrium' (MSE) experiment. In this, we mix the depleted solid phase obtained from the

SGGMP - Dunedin 2007

79 DPM experiment, with the synthetic glass that represents the partial melt, in the same proportions as predicted by the numerical model. We conducted the experiments between 925-1 OOO^C at 1.4 GPa; the same conditions as the earher set of DPM experiments on the same mafic amphibolitic starting material. The MSE experimental results show distinct differences in the melt and solid phase compositions and solid phase stability when compared with the results from the DPM experiments. The resulting melt compositions in the MSE experiments are lower in the An component (Fig. 1) and have higher Mg-numbers (Fig. 2) when compared with the DPM results. Modally, the charges have changed too, with a significant reduction in hornblende and plagioclase and an increase in garnet and clinopyroxene as a function of increasing temperature.

The results suggest that if dynamic melt segregation and equilibrium processes are active, they may modify the normally robust geochemical indicators, such as Mg-numbers, which are typically used to develop models of TTG petrogenesis.

Application to the Separation Point Batholith and other natural TTG compositions We have used as our amphibolitic starting material a natural mafic dike sample from Selwyn Creek, Fiordland. Figure 1 shows the experimental data from both the DPM and the MSE experiments and the natural compositions from the Separation Point Batholith, New Zealand. The MSE experimental data show a better overlap with the natural data suggesting that melt segregation processes may have influenced some compositional variation in the batholith.

Figure 2 shows experimental data compared with TTG sets compiled by Condie (2005) and Martin and Moyen (2002). The MSE experiments are higher in Mg# than the DPM experiments. The stability of hornblende is reduced in the MSE experiments and this may be the cause of the obersved increase in Mg# in the meh compositions.

SGGMP - Dunedin 2007

Figure 1: Ab-An-Or diagram plotting the direct partial melting results (diamonds) and melt segregation equilibrium results (circles, squares) showing the decrease in An component in the melt segregation equilibrium experiments. The crosses represent compositions from the Separation Point Batholith, New Zealand (Tulloch and Kimbrough, 2003)

Figure 2. Experimental glasses Mg# vs SiOi from meh segregation equilibrium experiments (small symbols), direct partial melting experiments (Price, 2005; Rapp et al, 1991; larger squares, white diamonds) and natural TTG data (largest squares and circle symbols). Direct partial melting experiments performed at pressures between 1.4 and 1.6 GPa are lower in Mg# than either the melt segregation equilibrium experiments performed or the natural TTGs.

SGGMP - Dunedin 2007

81 REFERENCES Condie, K.C., 2005. TTGs and adakites: are they both slab melts? Lithos 80; 33-44. Jackson, M.D., Gallagher, K., Petford, N., and Cheadle, M. J., 2005. Towards a coupled physical and chemical model for tonalite-trondhjemite-granodiorite magma formation. Lithos 79: 43-60. Martin, H., 1999. Adakitic magmas: Modem analogues of Archean granitoids. Lithos 46: 411-429. Martin, H., and Moyen, J-F., 2002. Secular changes in tonalite-trondhjemite-granodiorite composition as markers of the progressive cooling of Earth. Geology 30: 319- 322. Price, R. P. W., 2004. Testing the partial melting of a basaltic underplate: origin of Cretaceous granitoids in Fiordland, New Zealand. Unpublished MSc., University of Vermont. Rapp, R.P., Watson, E.B., Miller, C.F., 1991. Partial melting of amphibolite/eclogite and the origin of Archean trondhjemites and tonalites. Precambrian Research 51:1 -25. Rollinson, H., 2006. In: Evolution and Differentiation of the Continental Crust (Brown, and Rushmer, eds.), Cambridge University Press: pp 430-454. Tulloch, A.J., and Kimbrough, D.L., 2003. Paired plutonic belts in convergent margin and the development of high Sr/Y magmatism: The Peninsular Ranges Batholith of California and the Median Batholith of New Zealand. Geological Society of America, Special Paper 374.

Dating of fabric development by U-Pb isotopes: case studies from Fiordland, New Zealand

J.M. Scott, J.M. Palin, A.F. Cooper

University of Otago, Geology Department, Dunedin, New Zealand

The Gondwana margin rocks of Fiordland are regularly used as a natural laboratory for understanding the processes that take place during metamorphism, continental accretion, arc magmatism and continental break-up (e.g., Gibson et al, 1988; Ireland and Gibson, 1998; Daczko et al, 2001, 2002; Klepeis et al, 2004; Scott and Cooper, 2006). Superb rock exposure in parts of western and eastern Fiordland, coupled with U-Pb isotope ratios in zircon and/or titanite, enables determination of (1) development of metamorphic fabrics in both the Paleozoic and Mesozoic, (2) Cretaceous deformation associated with convergent margin tectonics, (3) Cretaceous extensional ductile deformation during rifting of New Zealand from Gondwana.

Paleozoic and Mesozoic metamorphic fabric development Detrital zircon spectra from metamorphosed sediments in western Fiordland indicate derivation from rocks formed during Gondwanan Precambrian orogenic episodes (Gibson and Ireland, 1996; Ireland and Gibson, 1998; Hollis et al, 2004). These sediments were metamorphosed in the Paleozoic or Mesozoic, or both (Ireland and Gibson, 1998). The extent of individual metamorphic events is, however, difficult to determine because of the intrusion of voluminous Early Cretaceous dioritic and gabbroic plutons, disruption of the metamorphic pile by extensional ductile shear zones and similarities in metamorphic conditions for both Mesozoic and Paleozoic events.

U-Pb ages, coupled with cathodoluminescence imaging, from the rims of metamorphosed igneous and detrital zircon grains provide a key to establishing the extent of Paleozoic and Mesozoic metamorphism. Bright rims with low Th/U ratios to detrital zircon from psammitic 82

83 schist in the Jaquiery River catchment (western Fiordland) demonstrate that the dominant fabric here in metasediments is Paleozoic (-350 Ma) (Fig. 1). Zircon grains from a kyanitebearing metasedimentary schist enclave in meta-anorthosite near George Sound (western Fiordland) also provide evidence for (complex) metamorphism in the Paleozoic (-350-340 Ma); however, low Th/U zircon rims reveal metamorphic growth in the Mesozoic (-116 Ma). Phase relationships and quantitative geothermobarometry imply that the Early Cretaceous zircon rims and metamorphic fabric probably formed during an up-pressure event.

Fabric development during convergent tectonism The Grebe Shear Zone is a sub-vertical mylonitic structure that separates the volcanic arc assemblage of the Median Tectonic Zone/ Median Batholith from the Early Paleozoic Gondwana margin metasediments of Western Fiordland (Fig. 2). U-Pb zircon intrusion ages of two shear zone tectonites (128 and 160 Ma) places an upper age to ductile deformation, while an undeformed West Arm Leucogranite apophysis (117 Ma) that cuts mylonitic rocks places a lower age. Zircon and titanite ages from structurally controlled syn-kinematic plutons (Refrigerator Orthogneiss and Puteketeke Granite), however, indicate that this zone of shear facilitated the transport of granitoid magma though the crust at 120 Ma. Because undeformed West Arm Leucogranite cuts across foliation in the Refrigerator Orthogneiss, fabric development must have occurred between orthogneiss protolith emplacement and leucogranite diking. Therefore, time constraints to volcanic arc juxtaposition, pluton emplacement and fabric development can be precisely determined through U-Pb data.

Fabric development during extensional tectonism The Mt. Irene Shear Zone (MISZ) in central Fiordland is a low angle extensional calcmylonite shear zone that juxtaposes Early Cretaceous dioritic orthogneiss footwall against a hanging-wall package of metamorphosed (Palaeozoic?) sediments (Fig. 3). Both the hangingwall and footwall are intruded by a suite of partially deformed granitoid dikes. Because dikes are both sheared within and cut across shear zone mylonitic fabric, this suite must be synkinematic. U-Pb zircon and titanite ages from one dike indicate emplacement at 108 Ma. This age requires that (1) shear zone fabric formed during and prior to this time, (2) that the MISZ is the oldest Early Cretaceous extensional shear zone in Fiordland and is a fore-runner to separation of New Zealand and Australia, (3) convergent tectonism (i.e., associated with the

Grebe Shear Zone) may have ceased by this time.

REFERENCES Daczko, N.R., Klepeis, K.A., and Clarke, G.L., 2001. Evidence of Early Cretaceous collisional-style orogenesis in northern Fiordland, New Zealand and its effects on the evolution of the lov^er crust. Journal of Structural Geology 23: 673-713. Daczko, N.R., Klepeis, K.A., and Clarke, G.L., 2002. Thermochronological evolution of the crust during convergence and deep crustal pluton emplacement in the Western Province of Fiordland, New Zealand. Tectonics 21: 1-17. Gibson, G.M., and Ireland, T.R., 1996. Extension of the Delamerian (Ross) Orogen into westem New Zealand; evidence from zircon ages and implications for crustal growth along the Pacific margin of Gondwana. Geological Society of America Bulletin 24: 1087-1090. Gibson, G.M., McDougall, I., and Ireland, T.R., 1988. Age constraints on metamorphism and the development of a metamorphic core complex in Fiordland, southern New Zealand. Geology 16: 405-408. Hollis, J.A., Clarke, G.L., Klepeis, K.A., Daczko, N.R., and Ireland, T.R., 2004. The regional significance of Cretaceous magmatism and metamorphism in Fiordland, New Zealand, from U-Pb zircon geochronology. Journal of Metamorphic Geology 22: 607627. Ireland, T.R., and Gibson, G.M., 1998. SHRIMP monazite and zircon geochronology of highgrade metamorphism in New Zealand. Journal of Metamorphic Geology 16:149-167. Klepeis, .K.A., Clarke, G.L., Gehrels, G., and Vervoort, J., 2004. Processes controlling vertical coupling and decoupling between the upper and lower crust of orogens: results from Fiordland, New Zealand. Journal of Structural Geology 26: 765-791. Scott, J.M., and Cooper, A.F., 2006. Early Cretaceous extensional exhumation of the lower crust of a magmatic arc: Evidence from the Mount Irene Shear Zone, Fiordland, New Zealand. Tectonics 25: TC3018.

Figure 1. Cathodoluminescence image of detrital zircon grains with Paleozoic metamorphic rims. Jaquiery River Catchment.

Psammite invaded by granite. Foliation in psammite is subparallel to shear zone

Massive, garnet-bearing, . biotite granite (OU 75687).

Mylonitic, porphyroclastic potassium feldspar - bearing granitoid (OU 75592). F= 010/86E, 004/70W, L=64/192. Cut by massive granitoid

\ Weakly foliated, pink porphyroclastic granite. F=015/85W

Zone of brittle faulting and cataclasis. Locally truncates eastern margin of mylonites Mylonitic quartz diorite (OU 75593). ^100 m

Figure. 2 Field sketch of the Grebe Shear Zone.

Mt Irene SpiayoftheMlSZ

Figure. 3. Cross-section through the Mt. Irene area. The Mt. Irene Shear Zone (light grey horizon) separates hanging-wall metasediments (dark grey) from footwall dioritic orthogneiss (white). Thick black lines are syn-kinematic granite dikes.

Oxygen isotopes in gem corundums, eastern Australia: further clues to their lithologic sources

F. Lin Sutherland \ G. Giuliani ^ A. E. Fallick ^ and G. B. Webb

' Geoscience, Australian Museum, Sydney, NSW. ^ Centre National des Recherche Scientific, Vandoeure-les Nancy, France ^ Scottish Universities Environmental Research Centre, East Kilbride, Scotland

Introduction The use of oxygen isotopes as a tracer in designating and interpreting the origin of gem corundums is now established as an important investigative tool for characterising source conditions, which include magmatic, metasomatic and metamorphic lithologies (Giuliani et al, 2005). Australian gem corundums are largely found in placer deposits, as xenocrysts derived from Late Cretaceous to Cenozoic basaltic eruptives (Sutherland and Webb, 2007). The deposits are part of an extensive gem corundum-basalt association that extends along the west Pacific continental margin from New Zealand in the south (Kiefert et al, 2006), through eastern Australia and then along eastern Asia into eastern Russia (Sutherland et al, 2004). They include polygenetic magmatic, metasomatic and metamorphic corundums (Graham et al, 2004). The Australasian suites include rubies, particularly in New South Wales deposits (Webb, 2007).

This study presents oxygen isotope data from a range of eastern Australian gem corundums, to provide a broad perspective on their oxygen isotope characteristics in relation to potential lithological sources. It adds to earlier oxygen isotope studies on some NSW and Tasmanian suites (Giuliani et al, 2005; Khin Zaw et al, 2006). It allows some tentative assignments to the sources that contributed corundums into basaltic magmas that erupted through the folded Neoproterozoic to Palaeozoic basements in the Kanmantoo, Lachlan, Thomson and New England fold belts (Veevers, 2000).

Materials and Analytical Methods A range of corundum grains between 2 to 10 mm across were selected to give a representative range of colours from ten separate localities, which extend from the central Queensland gem fields into a broad coverage of New South Wales gem fields (Table 1). Oxygen isotope analyses on individual grains used a laser fluorination method (Sharp, 1990), as modified by Macauley et al (2000). The fluorinating reagent was CIF3, and with a high power C02-laser there is no evidence for a grain size effect, nor is any correction factor necessary for comparison with conventional fluorination using external resistance fumaces (see Macauley et al, 2000). Precision and accuracy for isotopic homogenised material are ± 0.2 at 1 a and all data are reported as

relative to VSMO.

Results The

%o value for each grain from the separate localities is listed along with the

corundum colour and source type deduced by trace elements and mineral inclusion studies on related grains (Table 1). The Tasmanian sapphire values from a different study are also listed for comparison. The corundum isotope values range from 3.6 to 6.5 %o, but most results lie between 4 . 1 - 5 . 9 % o . The highest results ( 6 . 2 - 6 . 5 % o ) come from some Inverell, Barrington and Tumbarumba sapphires, which include some metamorphic types.

The great bulk of

magmatic sapphires fall in the range 4 . 1 - 6 . 2 % o , which includes the Tasmanian sapphires and is typical of results for such suites elsewhere (Giuliani et al,

2005).

89 Table 1: East Australian corundum suites, used for

comparative measurements

Field (Lat S, Long E)

Locality (workings)

Corundum colours (variation)

Origin type

Ref

Value

(State)

Rubyvale area (unlocalised)

Light blue Green Yellow Dark blue/green Light blue/green Light green/blue Green/yellow Green with blue Orange with yellow Grey blue Yellow Blue/yellow Light green Pink Light red Pale to dark blues; green; yellow. Pale-dark pink red; 'orange' Pink Deep pink Reddish pink Blue Green blue Blue/yellow Light blue/green Light blue Pinkish green^lue Pale to dark blues; white

Mag.

1,2

(Qld)

Anakie (23^25' 147^40')

4.9 4.5 5.7 5.3 4.8 5.1 4.8 5.1 4.7 6.5^ 6.5^ 5.3 5.4 3.9 3.6 4.65.8^ 5.16.2' 5.4 5.3 4.05 5.7 4.6 4.8 5.9 6.4 5.5 4.16.2^

D40948 (NSW) D4198 (NSW) D44353 (NSW) S.1-2 (NSW) YAR (NSW)

New England (29°50' 15riO')

Yarrowitch (31°30' 15r55')

Sapphire (Mary Anne Gulley) Nullamanna (Frazer Creek) Inverell (Dominion Leases) Inverell (unlocalised) South Yarrowich (S. Fenwicks Creek)

BAR (NSW)

Barrington (32 W 15r30')

Upper Mannmg River (Gummi Flats)

MR (NSW)

Cudgegong (33°35' 149° 0')

S. Wellington (Macquarie River)

D52712 (NSW) D5381 (NSW)

Oberon (33°60' 149°46')

E. Black Springs (Vulcan State Forest)

Tumbarumba (35°50' 148^00')

S. Laurel Hill (Tumbarumba Creek)

NETl-12 Weldborough (Tas) (41°15' 1 4 8 W )

Weld River (Ringarooma River)

(%o)

it.

((

Mag

1,2,3 ((

Int

Mag.

Mag

1,4

Met.

a u

Mag.

3,5,6

Met u

3,5,6 a

Met.

1,2,7

Mag.

1,2

((

Mag. Met. Int. Mag. u

1,2,8 a

2,5

Mag., magmatic. Met., metamorphic. Int., Intermediate. 1 This study. 2 Sutherland and Schwarz2001. 3 Giuliani

a/., 2005). 4 Sutherland

a/., 2005. 7 Khin Zaw er a/., 2006.

8 Roberts er a/., 2004. 9 Sutherland

a/., 2003. 10 Sutherland er a/., 2002. ® From on-line

data repository, Giuliani et al, (2005);

'' From Khin Zaw et al., 2006.

Discussion The East Australian corundum O isotope values are compared with those for corundums from a range of lithologies in Figure 1. The dominant Australian magmatic sapphire values (4-6 %o)

fall in the range for corundums in syenites ( 4 . 8 - 8 . 3 %o; Fig.l, G. Giuliani data). The

lower end values may include sapphires that crystallised from melts in equilibrium with ultramafic assemblages, even within mantle settings as demonstrated in the French Pyrenees (Pin et al, 2 0 0 6 ; G. Giuliani data). The higher end Australian magmatic sapphire values probably represent contamination of the parent syenitic melts with more silica-rich crustal

90 components.

Ruby suites from Yarrowitch and Cudgegong show low values that fall within and partly within values for rabies from mafic gneisses and are probably related to ultramafic to mafic granulite hosts (Sutherland et al, 2003). The Barrington and some Cudgegong rubies show higher values that fall within the range of desilicated

felsic/ultramafic associations

(plumasites).

The East Australian results indicate that considerable local variations occur in the magmatic sapphire values within the main Central Queensland and New England gem fields and also between the different gem fields. This suggests diverse magmatic as well as a range of metamorphic/ metasomatic sources contributed corundum xenocrysts into passing basaltic magmas along the length of the volcanic field.

Acknowledgements

Australian Museum, Ian Graham and associates for field and collection support.

REFERENCES Giuliani, G., Fallick, A.E., Gamier, V., France-Lanord, C., Ohenstteter, D., and Schwarz, D., 2005. Oxygen isotope composition as a tracer for the origins of rubies and sapphires. Geology 33: 249-252. Graham, I.T., Sutherland, F.L., Webb, G., and Fanning, C.M., 2004. Polygenetic corundums from New South Wales gem fields, Australia. In: Metallogeny of the Pacific Northwest: tectonics, magmatism and metallogeny of active continental margins (A.I. Khanchuk et al., eds.). Dalnauka, Vladivostok: 336-339. Khin Zaw, Sutherland, F.L., Dellapasqua, F., Ryan, C.G., Tzen-Fui, Y., Menargh, T.P., and Duncan, D., 2006. Contrasts in gem corundum characteristics, eastem Australian basalt fields: Trace elements, fluid/melt inclusions and oxygen isotopes. Mineralogical Magazine 70: 669-687. Kiefert, L., Krzemnicki, M.S., Du Toit, G., Befi, R., and Schmetzer, K., 2006. Sapphires

91 from New Zealand. Gems and Gemology 42: 113-114. Macaulay, C.I., Fallick, A.E., Hazeldine, R.S., and Graham, C.M., 2000. Methods of laserbased stable isotope measurement applied to diagenetic cements and hydrocarbon reservoir quality. Clay Minerals 35: 313-322. Pin, C., Monchoux, P., Paquette, J.-L., Azambre, B., Wang, R.C., and Martin, R.F., 2006. Igneous albitite dikes in orogenic Iherzolites, western Pyrenees, France: a possible source for corundum and alkali feldspar xenocrysts in basaltic terrains II. Geochemical and petrogenetic considerations. Canadian Mineralogist 44: 843-856. Roberts, D.L., Sutherland, F.L., Hollis, J.D., Kennewell, P., and Graham, I.T., 2004. Gemstone characteristics, North-East Barrington Plateau, NSW. Journal and Proceedings of the Royal Society of New South Wales 137: 99-122. Sharp, Z.D., 1990. A laser-based microanalytical method for the in situ determination of oxygen isotope ratios in silicates and oxides. Geochemica et Cosmochimica Acta 54: 1353-1357. Sutherland, F.L., and Schwarz, D., 2001. Origin of gem corundum from basaltic fields. Australian Gemmologist 21: 30-33. Sutherland, L., and Webb, G., 2007. Australian sapphires and rubies. Rocks and Minerals 82:116-125. Sutherland, F.L., Graham, I.T., and Webb, G., 2004. Sapphire-ruby-zircon deposits from basalt fields, west Pacific continental margins. In: Metallogeny of the Pacific Northwest: tectonics, magmatism and metallogeny of active continental margins (A.I. Khanchuk et al., eds.). Dalnauka, Vladivostok: 385-387 Sutherland, F.L., Coenraads, R.R., Schwarz, D., Raynor, L.R., Barron, B.J., and Webb, G.B., 2003. Al-rich diopside in alluvial ruby and corundum-bearing xenoliths, Australian and SE Asian basah fields. Mineralogical Magazine 67: 717-732. Sutherland, F.L., Graham, I.T., Pogson, R.E., Schwarz, D., Webb, G.B., Coenraads, R.R., Fanning, C.M., Hollis, J.D., and Allen, T.C., 2002. The Tumbarumba basaltic gem field. New South Wales: in relation to sapphire-ruby deposits in eastern Australia. Records of the Australian Museum 54: 215-248. Sutherland, F.L., Graham, I.T., Webb, G.B., Pogson, R.E., Giuliani, G. and Fallick, A.E., 2005. New sapphire-ruby sources, Yarrowitch basalt field, eastern New South Wales. 91

In: SGGMP-Port Macquarie (Graham, I.T., ed.). Geological Society of Australia Abstract Series 76: 133-136. Veevers, J.J., 2000. Billion-year earth history of Australia and neighbours in Gondwana. Gemoc Press. Webb, G., 2007. Ruby suites from New South Wales. Austrahan Gemmologist 23: 99-117.

51^0 corundum (%o, V-SMOW)

Figure 1. Oxygen isotope ranges, Australian gem corundums and potential corundum-source rocks

U-Pb geochronology of mid-Paleozoic granitic magmatism in New Zealand

Andy Tulloch', Jahandar Ramezani^ David Kimbrough^ Kevin Faure"^ and Andrew Allibone^

'GNS Science, Private Bag 1930, Dunedin ^ Earth, Atmospheric and Planetary Sciences, MIT, USA ^ Dept of Geological Sciences, San Diego State University, USA GNS Science, PO Box 30368, Lower Hutt ^ Rodinian Pty Ltd, P.O. Box 1804, Fyshwick ACT2609, Australia E-mail: a.tulloch@gns.cri.nz

Introduction Mid-Paleozoic granites are widespread in formerly contiguous parts of Gondwana - the Lachlan Fold Belt (LFB) and the New England Fold Belt (NEFB) of eastern Australia, North Victoria Land and Marie Byrd Land in Antarctica. Previous workers in New Zealand reported a major episode at 380-370 Ma and a smaller event at 330 Ma (Muir et al 1994, 1996). We have recognised and mapped 6 suites of mid-Paleozoic granites in western New Zealand; two S-types, two I-types and two A-type suites.

We have dated 27 plutons representing these suites using ID-TIMS. Most ages were determined on single zircon crystals at MIT by low-blank (mostly <lpg Pb) high precision (0.05-0.1% internal errors) analyses that allow robust assessment of

- ^"'Pb/^^^U

concordancy, and thus accuracy. Careful selection of high-quality melt precipitated grains from 99% pure mineral separates obtained from carefully examined and processed samples was successful in avoiding inheritance in > 75%) of analyses of zircon from S-type granites. Pb-loss was minimised by either standard air abrasion techniques, or the newand more

94 efficient thermal annealing-acid leaching (CA-TIMs) technique of Mattinson (2005). Examples will be presented to illustrate apparent open system behaviour indicated by discordance.

Age results Results reveal episodic emplacement of granitic rocks over ~ 80 Ma, with steadily declining volumes from a large burst in the Late Devonian (-370 - 360 Ma) to a smaller burst in the Early Carboniferous (355 - 340 Ma) and very minor, mostly A-type magmatism, at 320 - 305 Ma. The S-type suites are partly coeval with the Paringa (369 - 361 Ma) and Tobin (352 - 342 Ma) I-type suites, respectively. Age and S-type character decrease from 370 - 342 Ma eastwards (trenchwards) over ~ 260 km normal to the Gondwana margin. S-type granites of Karamea (370 - 368, 350 - 354 Ma) and Ridge (« 354 - 345 Ma) suites are essentially restricted to the Buller and Takaka Early Paleozoic metasedimentary terranes respectively, suggesting significant influence of these sedimentary terranes on granite composition. A likely minimum age for Buller/Takaka terrane amalgamation is given by appearance of Paringa Suite (369 - 361 Ma) in both terranes. The boundary is stitched by a Ridge suite pluton at 355 Ma.

Correlation with Australia The New Zealand record is dominated by -370 - 340 Ma granites which are 30 - 40 Ma younger than the bulk of the LFB granitoids. In the LFB the 384 - 350 Ma Central Victorian Magmatic Province (VandenBurg et al, 2000; essentially the Melbourne Terrane of Chappell et al, 1988), and the 360 - 320 Ma Bathurst Terrane (Chappell, 1994) are coeval with -370 340 Ma granitic magmatism in New Zealand. Conversely, most New Zealand Palaeozoic granites are distinctly older than the bulk of NEFB granites

360 - 280 Ma). In addition,

New Zealand S-type plutons extend to less radiogenic compositions (Average Srj = 0.7082 for Karamea Suite, 0.7059 for Ridge Suite) than the archetypal S-types of the LFB, but are more radiogenic than most of the Hillgrove Suite of the NEFB (Flood and Shaw, 1977; Landenberger and Collins, 1995). Thus, it appears the New Zealand sector may represent a record of magmatism in a space and time interval that bridges the bulk of magmatism in the two Australian regions. Together, these three parts of the Paleozoic Gondwana margin provide a more complete record of continental margin growth and development, spanning

95 some 150 Ma.

The Karamea suite Paleozoic magmatism in New Zealand is overwhelmingly dominated by the Karamea Suite S-type granites, which form much of the Karamea Batholith and a number of outlying plutons. New high-precision age data for Karamea Suite S-type plutons suggests that this suite was emplaced in a very rapid burst, perhaps as little as 2 Ma. This rate of Karamea Suite magma generation ( - 3 8 W / k m arc/Ma; conservatively assuming 5 km thickness) by partial melting of the lower crust would have required a much greater thermal flux from the mantle than hitherto realised, similar to the magma eruption rate of the current Taupo Volcanic Zone (~30km^/km arc/Ma). An extensional tectonic setting, allowing rapid influx of asthenospheric magmas into the crust, appears likely. Sr-Nd, '^O isotopes and zircon inheritance suggest major components of recycled sedimentary crust and possibly mafic lower crust, but a significant juvenile manfle component, and thus crustal growth, is again indicated.

Given the indicated high thermal flux into the lower crust it is likely that the upper crust in western New Zealand was also affected by high heat flow and hydrothermal alteration. We suggest that the Karamea event is thus a good candidate for developing Taupo-scale hydrothermal systems which, speculatively, may have led to at least some of the Paleozoic gold mineralisation (eg. Bierlein et al, 2004) in Westland.

REFERENCES Bierlein, P.P., Christie, A.B., and Smith, P.K., 2004. A comparison of orogenic gold mineralisation in central Victoria (AUS), western South Island (NZ) and Nova Scotia (CAN): implications for variations in the endowment of Palaeozoic metamorphic terrains. Ore Geology Reviews 25: 125. Chappell, B.W., 1994. Lachlan and New England: Fold Belts of contrasting Magmatic and Tectonic development: Journal and Proceedings Royal Society of New South Wales 127: 47-59. Chappell, B.W., White, A.J.R., and Hine, R., 1988. Granite provinces and basement terranes in the Lachlan Fold Belt, southeastern Australia: Australian Journal of Earth Sciences 95

96 35:501-521. Flood, R.H., and Shaw, S.E., 1977. Two "S-Type" granite Suites with low initial ^^Sr/^^Sr ratios from the New England Batholith, Australia. Contributions to Mineralogy and Petrology 61: 163-173. Landenberger, B., and Collins, W.J., 1995. S-type granites of the Hillgrove Plutonic Suite, Eastern Australia: products of partail melting of an intermediate greywacke source. United States Geological Survey Circular 1129 (Hutton Symposium III): 86. Mattinson, J.M., 2005. Zircon U-Pb chemical abrasion ("CA-TIMS") method: Combined annealing and multi-step partial dissolution analysis for improved precision and accuracy of zircon ages. Chemical Geology 220: 47-56. Muir, R.J., Ireland, T.R., Weaver, S.D., and Bradshaw, J.D., 1994. Ion microprobe U-Pb zircon geochronology of graniticmagmatism in the Western Province of the South Island, New Zealand. Chemical Geology 113: 171-189. Muir, R.J., Ireland, T.R., Weaver, S.D., and Bradshaw, J.D., 1996. Ion microprobe dating of Paleozoic granitoids: Devonian magmatism in New Zealand and correlations with Australia and Antarctica. Chemical Geology 127: 191-210. VandenBerg, A.H.M., Willman, C.E., Maher, S., Simons, B.A., Cayley, R.A., Taylor, D.H., Morand, V.J., Moore, D.H., and Radojkovic, A., 2000. The Tasman Fold Belt system in Victoria. Geological Survey of Victoria.

SOUTH ISLAND, NEW ZEALAND 167^E Nelson Western Province

^ /

Fiord l a n d / ^ ^

Eastern Province —•• Stewart Lsland Anatok! Fault EASTERN^ PROVINCE

50 km

Paleozoic granite suites

1 g Karamea (K) S-type i f Ridge l-type

•

Paringa Tobin (T)

A-type

' i Foulwind (F) granites r undifferentiated Small bodies Ie.rra..ng..s

—

Old Quarry Fault

STEWART ISLAND Takaka Terrane

Buller, Ordovician Takaka, CambrianDevonian

VA/estern Province

New constraints from Tonga-Kermadec on the origin of O-Hf-Os isotope signatures in oceanic arc lavas

Simon Turner^ Monica Handler^, Ilya Bindeman^

GEMOC, Department of Earth and Planetary Sciences, Macquarie University, Sydney NSW 2109, Australia ^ School of Geography, Environment and Earth Sciences, Victoria University of Wellington, PO Box 600, New Zealand ^ Department of Geology, University of Oregon, Eugene, OR 97403, USA

O, Hf and Os isotope ratios have been measured on selected lavas from the Tonga-Kermadec arc that have previously been analysed for Be isotopes which provides an unambiguous tracer of subducted sediment. O isotope values overlap with those of MORB and indicate minimal interaction with the arc crust. O isotopes do not increase northwards with increasing subduction rate as would be expected if slab derived fluids had high O isotope ratios. The northward decrease in HFSE concentrations reflects prior depletion not increasing extents of melting. Hf isotopes are strongly negatively correlated with Be isotopes. Simple binary mixing of subducted pelagic sediment into the mantle wedge can replicate the Hf-Be isotope correlation, without recourse to fluid mobility of Hf However, the sediment component must have been stored in the mantle wedge for several Myr, as suggested previously. The same mixing model can replicate those samples with the lowest Os isotopes and one sample has subchondritic Os indicating that some melt depletion of the mantle wedge pre-dates the opening of the Lau Basin. Thus, some slab-derived Os isotope signatures are preserved which requires rapid magma ascent to minimise meh - wall rock interaction. Higher Os isotope ratios in some samples may reflect interaction with the arc crust. There is no overlap between the Hf isotope composition of the north Tonga arc lavas and either the Samoan plume or the subducting Louisville volcaniclastic sediments. Thus, the unusual Pb and Nd isotope signatures in these lavas must derive from the Louisville sediments but be added with minimal mobilisation of Hf indicating conservative behaviour for this element.